# SetGet: full tool and error reference > Every SetGet tool with its inputs and a worked example, then every error code. The short overview is at https://setget.net/llms.txt. ## Calling a tool `POST https://setget.net/v1/tools/` with `Content-Type: application/json` and body `{"input": {...}}`. A successful response is `{"tool": "", "ms": , "result": {...}}`. A failure is `{"error": {"code", "message", "fix", "retryable", "docs", ...}}`. ## File tools Tools with a "Files in, files out" section take file references instead of text. Upload a file with `curl -T FILE https://setget.net/drop/` (the reply has a line `ref /f1`) or `POST https://setget.net/v1/drops`, pass the reference as the input, and add `"wait": 30` to the request. The reply carries `result` and `files`; each result file has a `url` to download and a `ref` to pass to the next tool. A job that outlasts the wait answers 202 with `job`: read `GET https://setget.net/v1/jobs/?wait=30` until `status` is `done`, `failed` or `cancelled`. Files are public by link and expire, one hour by default. ## Images Convert, resize, compress, crop and inspect images and GIFs. Upload a file, get a file back. ### Add a text watermark Add a text watermark to an image in a corner, in the centre or tiled across it, with adjustable opacity and size. - Endpoint: `POST https://setget.net/v1/tools/image-watermark` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-watermark` - Page: https://setget.net/tools/image-watermark #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to watermark. - `text` (required, string): The watermark text, for example a name or a copyright line. Latin, Greek or Cyrillic letters, digits and punctuation. - `position` (optional, string, one of `top-left`, `top-right`, `bottom-left`, `bottom-right`, `center`, `tile`, default `"bottom-right"`): Where to put it. tile repeats it diagonally across the whole image. - `opacity` (optional, integer, default `50`, range 1..100): How solid the text is, from 1 (barely visible) to 100 (fully solid). - `size` (optional, number, default `5`, range 1..50): Height of the text as a percentage of the shorter side of the image. - `color` (optional, string, default `"#ffffff"`): Colour of the text. A hex colour such as #ff8800, or a name such as white or black. #### Example 1: Bottom-right corner Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","text":"(c) Example Studio"},"wait":30} ``` #### Example 2: Tiled across the image Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","text":"DRAFT","position":"tile","opacity":25,"size":8,"color":"#000000"},"wait":30} ``` **Why is the size relative?** So the watermark looks the same on a small preview and a large original. 5 means the letters are about 5% of the shorter side tall. **Can a watermark be removed?** A small corner mark can be cropped off. A tiled watermark across the image is much harder to remove. **Which characters can I use?** Latin letters including accents, Greek, Cyrillic, digits, punctuation and symbols such as the copyright sign. Chinese, Japanese, Korean, Arabic and emoji are not supported yet. **What if the text is too long?** It is made smaller until it fits inside the image; the result reports the text height used. ### Add padding or a border Add padding, margin or a coloured border around an image, the same on every side or set per side. - Endpoint: `POST https://setget.net/v1/tools/image-pad` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-pad` - Page: https://setget.net/tools/image-pad #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to pad. - `padding` (optional, integer, range 0..5000): Pixels to add on every side that is not set separately. - `top` (optional, integer, range 0..5000): Pixels to add above the image. - `right` (optional, integer, range 0..5000): Pixels to add to the right of the image. - `bottom` (optional, integer, range 0..5000): Pixels to add below the image. - `left` (optional, integer, range 0..5000): Pixels to add to the left of the image. - `color` (optional, string, default `"#ffffff"`): Colour of the added area. transparent works in formats that support it and becomes white in others, such as JPG. A hex colour such as #ff8800, or a name such as white or black. #### Example 1: 20 pixel white border Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","padding":20},"wait":30} ``` #### Example 2: Black bars above and below Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","top":40,"bottom":40,"color":"#000000"},"wait":30} ``` #### Example 3: Transparent margin Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1","padding":16,"color":"transparent"},"wait":30} ``` **How do I make an image square without cropping?** Add padding to the two short sides. For a 640 x 400 image, set top and bottom to 120 each to get 640 x 640. **Does the image itself change?** No. The picture keeps its size and pixels; the canvas around it grows. ### Add text to an image Write text on an image at a position you choose, with size, colour, an outline or a background box. Multi-line. - Endpoint: `POST https://setget.net/v1/tools/image-text-overlay` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-text-overlay` - Page: https://setget.net/tools/image-text-overlay #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to write on. - `text` (required, string): The text to draw. Use line breaks (or \n) for several lines. Latin, Greek or Cyrillic letters, digits and punctuation. - `x` (optional, integer, range 0..): Left edge of the text, in pixels from the left of the image. Leave empty to use position. - `y` (optional, integer, range 0..): Top edge of the text, in pixels from the top of the image. Leave empty to use position. - `position` (optional, string, one of `top-left`, `top`, `top-right`, `left`, `center`, `right`, `bottom-left`, `bottom`, `bottom-right`, default `"center"`): Where to put the text when x or y is not given. - `size` (optional, integer, default `32`, range 6..1000): Height of the text in pixels. - `color` (optional, string, default `"#ffffff"`): Colour of the text. Add two more hex digits for see-through text, such as #ffffff80. A hex colour such as #ff8800, or a name such as white or black. - `outline` (optional, string): Colour of a thin outline around the letters, which keeps text readable on busy pictures. Leave empty for none. - `box` (optional, string): Colour of a filled box behind the text, for example #00000080 for half-transparent black. Leave empty for none. #### Example 1: Centred text Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","text":"Hello world"},"wait":30} ``` #### Example 2: Two outlined lines at the bottom Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","text":"Top line\nSecond line","position":"bottom","size":28,"outline":"#000000"},"wait":30} ``` #### Example 3: On a dark box at a fixed spot Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","text":"SALE","x":20,"y":20,"size":48,"color":"#ffffff","box":"#000000cc"},"wait":30} ``` **How do I keep text readable on any picture?** Add an outline in a contrasting colour (white text with a black outline works on almost anything), or put a half-transparent box behind it. **Which characters can I use?** Latin letters including accents, Greek, Cyrillic, digits and punctuation. Chinese, Japanese, Korean, Arabic and emoji are not supported yet; the tool tells you which characters it cannot draw. **Can I choose the font?** Not yet. One clean sans-serif font is built in. **What if the text does not fit?** It is drawn at the size you asked for and cut off at the edge of the image. The result says when that happened, with the space the text needed. ### Adjust brightness, contrast and colour Adjust the brightness, contrast, saturation, sharpness and gamma of an image. 1 leaves a setting unchanged. - Endpoint: `POST https://setget.net/v1/tools/image-adjust` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-adjust` - Page: https://setget.net/tools/image-adjust #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to adjust. - `brightness` (optional, number, default `1.0`, range 0..4): Below 1 is darker, above 1 is brighter. 0 is black. - `contrast` (optional, number, default `1.0`, range 0..4): Below 1 is flatter, above 1 is punchier. 0 is solid grey. - `saturation` (optional, number, default `1.0`, range 0..4): Below 1 is more muted, above 1 is more vivid. 0 is black and white. - `sharpness` (optional, number, default `1.0`, range 0..4): Below 1 is softer, above 1 is sharper. 0 is blurred. - `gamma` (optional, number, default `1.0`, range 0.1..5): Lightens (above 1) or darkens (below 1) the middle tones without clipping black or white. #### Example 1: A little brighter and punchier Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","brightness":1.2,"contrast":1.1},"wait":30} ``` #### Example 2: Muted and sharper Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","saturation":0.5,"sharpness":1.5},"wait":30} ``` #### Example 3: Lift the shadows Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","gamma":1.8},"wait":30} ``` **What do the numbers mean?** 1 means no change. 1.2 is 20% more and 0.8 is 20% less. Small steps go a long way: most photos need values between 0.8 and 1.3. **Should I use brightness or gamma for a dark photo?** Gamma. Brightness lifts everything and can wash out bright areas; gamma lifts the shadows and middle tones while keeping white as white. **In what order are they applied?** Gamma, brightness, contrast, saturation, then sharpness. ### App icon pack for iOS and Android Make every iOS and Android app icon size as PNG files from one 1024 x 1024 image. Upload one square image of 1024 x 1024 pixels or more. iOS icons are made at 20, 29, 40, 58, 60, 76, 80, 87, 120, 152, 167, 180 and 1024 pixels (ios-20.png to ios-1024.png). Android launcher icons are made at 48 (mdpi), 72 (hdpi), 96 (xhdpi), 144 (xxhdpi) and 192 (xxxhdpi) pixels, plus the 512 pixel Play Store icon (android-mdpi-48.png to android-play-store-512.png). - Endpoint: `POST https://setget.net/v1/tools/app-icon-pack` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `app-icon-pack` - Page: https://setget.net/tools/app-icon-pack #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The icon artwork. Square and at least 1024 x 1024 pixels works best. - `platforms` (optional, string, one of `both`, `ios`, `android`, default `"both"`): Which icons to make. - `background` (optional, string, default `"#ffffff"`): Colour behind transparent areas, around a non-square image and in any padding. iOS icons are always solid: transparent becomes white there. - `padding` (optional, integer, default `0`, range 0..40): Empty space on every side, as a percentage of the icon size. #### Example 1: All iOS and Android sizes Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Android icons with transparency kept Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1","platforms":"android","background":"transparent"},"wait":30} ``` **Why are my iOS icons not transparent?** Apple rejects app icons with transparency, so iOS icons are always placed on the background colour (white if you asked for transparent). **What if my image is smaller than 1024 pixels?** The icons are still made, but the larger ones are enlarged and will look soft. The result carries a warning when this happens. **Are the corners rounded?** No. iOS and Android round the corners themselves, so upload square artwork. ### BlurHash of an image Compute the BlurHash string of an image: a short text placeholder that decodes to a blurred preview. - Endpoint: `POST https://setget.net/v1/tools/image-blurhash` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-blurhash` - Page: https://setget.net/tools/image-blurhash #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to encode. - `components_x` (optional, integer, default `4`, range 1..9): Detail across the image, 1-9. More gives a longer string with more detail. - `components_y` (optional, integer, default `3`, range 1..9): Detail down the image, 1-9. #### Example 1: Standard 4 x 3 BlurHash Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: More detail Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","components_x":6,"components_y":4},"wait":30} ``` **What is a BlurHash for?** Store the short string next to an image, and your app or page can show a soft blurred preview straight away while the real image loads. Decoders exist for most platforms. **Which component counts should I use?** 4 x 3 suits landscape images and 3 x 4 portrait ones. More components give more detail but a longer string: its length is 4 + 2 x components_x x components_y. **What happens to transparency?** Transparent areas are treated as white, because a BlurHash cannot hold transparency. ### Change GIF speed, reverse or boomerang Speed up or slow down an animated GIF, WebP or PNG, play it backwards, or make a forward-and-back boomerang. - Endpoint: `POST https://setget.net/v1/tools/gif-speed-reverse` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `gif-speed-reverse` - Page: https://setget.net/tools/gif-speed-reverse #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The animation to change: GIF, animated WebP or animated PNG. - `speed` (optional, number, default `1`, range 0.1..10): Speed factor: 2 plays twice as fast, 0.5 at half speed. - `reverse` (optional, boolean, default `false`): Play the frames in reverse order. - `boomerang` (optional, boolean, default `false`): Play forward and then back again, for a seamless loop. #### Example 1: Twice as fast Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","speed":2},"wait":30} ``` #### Example 2: Play backwards Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","reverse":true},"wait":30} ``` #### Example 3: Slow boomerang Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","speed":0.5,"boomerang":true},"wait":30} ``` **How fast can a GIF go?** Browsers do not show frames for less than 20 milliseconds, so frame times are never set below that. To go faster still, drop frames with gif-optimize. **What does boomerang do with reverse?** The animation is reversed first, then played back to its start: backward and then forward. **Is the format kept?** Yes. A GIF stays a GIF, a WebP stays a WebP and an animated PNG stays a PNG. ### Colour palette from an image Extract the dominant colours of an image as hex and RGB values with their share, plus a swatch image. - Endpoint: `POST https://setget.net/v1/tools/image-palette` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-palette` - Page: https://setget.net/tools/image-palette #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to take colours from. - `count` (optional, integer, default `5`, range 1..16): How many colours to find. - `swatch` (optional, boolean, default `true`): Also make a PNG strip showing the colours. #### Example 1: Five main colours Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Three colours, no swatch image Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","count":3,"swatch":false},"wait":30} ``` **How are the colours chosen?** Pixels are sorted into "count" groups of similar colour, and each group gives its average. An image with no more than "count" distinct colours returns exactly those colours. **What is the percent?** The share of the image that falls into that colour group. The values add up to 100. **Are transparent areas counted?** No. Pixels that are more than half transparent are ignored. ### Compress image Make an image file smaller by re-encoding it at a chosen quality, with an optional maximum width. - Endpoint: `POST https://setget.net/v1/tools/image-compress` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-compress` - Page: https://setget.net/tools/image-compress #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to compress. - `quality` (optional, integer, default `75`, range 1..100): From 1 (smallest file) to 100 (best quality). 75 is a good balance for photos. For PNG, values below 100 reduce the number of colours. - `max_width` (optional, integer, range 1..16000): Shrink the image to this width in pixels if it is wider. Smaller dimensions save the most space. - `format` (optional, string, one of `jpg`, `png`, `webp`, `avif`): Format of the result. Leave empty to keep JPG, PNG, WebP and AVIF as they are; other formats such as GIF, BMP and TIFF become WebP. #### Example 1: Compress a photo Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","quality":60},"wait":30} ``` #### Example 2: Smaller and narrower, as WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","quality":70,"max_width":320,"format":"webp"},"wait":30} ``` #### Example 3: Animated GIF to a smaller animated WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **Which quality should I pick?** Start at 75. Photos usually look the same down to about 60. Go to 85 or higher for images with text or sharp edges. **How is a PNG compressed?** At quality 100 it is repacked with no change to the pixels. Below 100 the colours are reduced to 256 or fewer, which often cuts the size by more than half and suits logos, screenshots and drawings. **Why did my file get bigger?** It was already compressed harder than the quality you chose. The result reports this; try a lower quality, a max_width, or WebP. **What saves the most?** Reducing the dimensions with max_width, then switching to WebP or AVIF. ### Compress image to a file size Compress an image to fit under a target file size in KB, such as 100 KB or 2 MB, for upload limits and forms. - Endpoint: `POST https://setget.net/v1/tools/image-compress-to-size` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-compress-to-size` - Page: https://setget.net/tools/image-compress-to-size #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to compress. - `target_kb` (required, number, range 0.1..50000): Largest allowed file size in kilobytes. 1 MB is 1000 KB. - `format` (optional, string, one of `jpg`, `webp`, `avif`): Format of the result. Leave empty to keep JPG, WebP or AVIF; other images become JPG, or WebP when they have transparency. #### Example 1: Fit a photo under 10 KB Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","target_kb":10},"wait":30} ``` #### Example 2: Under 4 KB as WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","target_kb":4,"format":"webp"},"wait":30} ``` **How does it reach the size?** It looks for the highest quality that fits. If even a low quality is too big, it makes the image smaller in pixels and tries again. The result reports the quality and scale used. **Is the result always under the target?** Yes. A kilobyte is counted as 1000 bytes, so the file also passes limits that count 1024 bytes per KB. If the target cannot be reached you get an error instead of an oversized file. **What happens to transparency?** WebP and AVIF keep it. JPG cannot, so transparent areas turn white. ### Contact sheet maker Lay out up to 100 images as a grid of thumbnails on one sheet, with optional file-name labels. - Endpoint: `POST https://setget.net/v1/tools/contact-sheet` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `contact-sheet` - Page: https://setget.net/tools/contact-sheet #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `images` (required, files, a list of file references such as `["DROP_ID/f1"]`, at most 100, accepts image): The images to show, up to 100, in order. - `columns` (optional, integer, default `4`, range 1..20): Thumbnails per row. - `thumb_size` (optional, integer, default `160`, range 32..512): Width and height of each thumbnail cell in pixels. Images are scaled to fit inside. - `labels` (optional, boolean, default `true`): Write each file name under its thumbnail. - `background` (optional, string, default `"#ffffff"`): Colour of the sheet. A hex colour, a colour name or transparent. - `format` (optional, string, one of `png`, `jpg`, `webp`, default `"png"`): File format of the sheet. #### Example 1: Four thumbnails with labels Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg), `DROP_3/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png), `DROP_4/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1","DROP_3/f1","DROP_4/f1"]},"wait":30} ``` #### Example 2: Dark sheet without labels Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1"],"columns":2,"thumb_size":96,"labels":false,"background":"#111827"},"wait":30} ``` **Are images cropped?** No. Each image is scaled down to fit inside its cell and keeps its shape. **What happens to long file names?** They are shortened with three dots so they fit under the thumbnail. **Which frame of a GIF is shown?** The first one. ### Convert animated images Convert an animation between GIF, animated WebP, animated PNG (APNG) and AVIF, keeping the frame timing. - Endpoint: `POST https://setget.net/v1/tools/animated-convert` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `animated-convert` - Page: https://setget.net/tools/animated-convert #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The animation to convert: GIF, animated WebP, animated PNG or AVIF. - `format` (optional, string, one of `gif`, `webp`, `png`, `avif`, default `"webp"`): Format to convert to. png makes an animated PNG. - `quality` (optional, integer, default `80`, range 1..100): Quality for WebP and AVIF, 1-100. Higher looks better and is larger. GIF and PNG ignore it. #### Example 1: GIF to animated WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: GIF to animated PNG Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","format":"png"},"wait":30} ``` #### Example 3: GIF to animated AVIF Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","format":"avif","quality":60},"wait":30} ``` **Which format is smallest?** Usually AVIF, then WebP. Both are far smaller than GIF for the same picture. Animated PNG is lossless and often the largest. **Is the timing kept?** Yes, every frame keeps its own duration. Frames shorter than 20 milliseconds are raised to 20, the shortest time browsers honour. **What about a still image?** It is converted as a single frame. For still images image-convert offers more formats. **Does converting to GIF lose quality?** It can: GIF holds 256 colours per frame and transparency is on or off, with no soft edges. ### Convert image format Convert an image between PNG, JPG, WebP, AVIF, GIF, BMP, TIFF and ICO. Animated GIF and WebP stay animated. - Endpoint: `POST https://setget.net/v1/tools/image-convert` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-convert` - Page: https://setget.net/tools/image-convert #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to convert. - `format` (required, string, one of `png`, `jpg`, `webp`, `avif`, `gif`, `bmp`, `tiff`, `ico`): The format to convert to. - `quality` (optional, integer, range 1..100): Quality from 1 (smallest file) to 100 (best) for JPG, WebP and AVIF. Leave empty for a sensible default. - `lossless` (optional, boolean, default `false`): Keep every pixel exact in WebP or AVIF, at the cost of a larger file. - `background` (optional, string, default `"#ffffff"`): Colour put behind transparent areas when the new format has no transparency (JPG and BMP). A hex colour such as #ff8800, or a name such as white or black. #### Example 1: PNG to WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","format":"webp"},"wait":30} ``` #### Example 2: Transparent PNG to JPG on white Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1","format":"jpg","background":"#ffffff"},"wait":30} ``` #### Example 3: Animated GIF to animated WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","format":"webp"},"wait":30} ``` **What happens to transparency?** PNG, WebP, AVIF, GIF, TIFF and ICO keep it. JPG and BMP cannot store it, so transparent areas are filled with the background colour, white unless you choose another. **Do animations survive?** Yes when the new format can hold one: GIF, WebP, PNG (as APNG) and AVIF. Other formats get the first frame, and the result says so. **How is an ICO made?** The image is fitted into a transparent square of at most 256 pixels and saved with the standard icon sizes up to that size. **Is camera data kept?** No. Camera details and location are dropped. The colour profile is kept, and a photo that was stored sideways is turned upright. ### Crop image Crop an image to a rectangle in pixels, or to an aspect ratio such as 16:9 or 1:1 around its centre or an edge. - Endpoint: `POST https://setget.net/v1/tools/image-crop` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-crop` - Page: https://setget.net/tools/image-crop #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to crop. - `x` (optional, integer, range 0..): Left edge of the crop, in pixels from the left of the image. Leave empty to place the crop by gravity. - `y` (optional, integer, range 0..): Top edge of the crop, in pixels from the top of the image. Leave empty to place the crop by gravity. - `width` (optional, integer, range 1..): Width of the crop in pixels. - `height` (optional, integer, range 1..): Height of the crop in pixels. - `aspect` (optional, string): Crop to this shape instead of a size, written width:height, for example 16:9 or 1:1. The largest area of that shape is kept. - `gravity` (optional, string, one of `center`, `top`, `bottom`, `left`, `right`, default `"center"`): Which part to keep when x or y is not given. #### Example 1: Cut out a rectangle Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","x":40,"y":20,"width":200,"height":120},"wait":30} ``` #### Example 2: Largest centred square Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","aspect":"1:1"},"wait":30} ``` #### Example 3: 16:9 from the top Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","aspect":"16:9","gravity":"top"},"wait":30} ``` **How do I find the pixel positions?** Open the image in any editor and hover over the corner you want: most show the x and y under the pointer. x counts from the left edge and y from the top, starting at 0. **Can I crop to a size without giving x and y?** Yes. Give width and height only and the crop is placed by gravity: the centre by default, or against the top, bottom, left or right edge. **Does cropping reduce quality?** PNG, GIF and other exact formats lose nothing. A JPG is saved again at high quality, which changes it very slightly. ### Favicon pack Make a complete favicon set from one image: favicon.ico, PNG icons, Apple touch icon and the HTML tags. Upload one image, ideally square and at least 512 x 512 pixels. You get favicon.ico (holding 16, 32 and 48 pixel versions), favicon-16.png, favicon-32.png, apple-touch-icon.png (180), icon-192.png and icon-512.png. The result also holds the HTML link tags for your page head and a site.webmanifest to save next to the icons. - Endpoint: `POST https://setget.net/v1/tools/favicon-pack` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `favicon-pack` - Page: https://setget.net/tools/favicon-pack #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to turn into icons. Square and at least 512 x 512 pixels works best. - `background` (optional, string, default `"transparent"`): Colour that fills the square around a non-square image and any padding. A hex colour, a colour name or transparent. - `padding` (optional, integer, default `0`, range 0..40): Empty space on every side, as a percentage of the icon size. - `site_name` (optional, string, default `"My site"`): Name of the site or app, used in the web manifest. - `theme_color` (optional, string, default `"#ffffff"`): Theme colour for the web manifest, as a hex colour. #### Example 1: Favicon set from a square logo Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Pad a wide image onto a dark square Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","background":"#0f172a","padding":10,"site_name":"Example"},"wait":30} ``` **Where do the files go?** Put all six files in the top folder of your site, save the manifest text as site.webmanifest next to them, and paste the HTML tags into the head of your pages. **Why is the Apple touch icon not transparent?** iPhones and iPads show transparent areas as black. When the background is transparent, the Apple icon alone is placed on white instead. **My image is not square. What happens?** It is centred on a square and the rest is filled with the background colour. Nothing is cropped or stretched. ### Grayscale image Turn a colour image into grayscale (black and white), or give it a warm sepia tone. Transparency is kept. - Endpoint: `POST https://setget.net/v1/tools/image-grayscale` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-grayscale` - Page: https://setget.net/tools/image-grayscale #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to convert to grayscale. - `sepia` (optional, boolean, default `false`): Give the result a warm brown tone, like an old photograph, instead of neutral grey. #### Example 1: Black and white Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Sepia tone Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","sepia":true},"wait":30} ``` **Is this pure black and white?** No. It is shades of grey, as in a black-and-white photograph. Each pixel keeps its brightness and loses its colour. **Does it make the file smaller?** Usually a little, because one grey value is stored per pixel instead of three colours. ### Image diff Compare two images pixel by pixel: get a highlighted difference image, the share of changed pixels and where they are. - Endpoint: `POST https://setget.net/v1/tools/image-diff` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-diff` - Page: https://setget.net/tools/image-diff #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `before` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The original image. - `after` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The changed image. - `tolerance` (optional, integer, default `0`, range 0..255): How much a colour value (0-255) may differ before a pixel counts as different. 0 counts every change. - `resize` (optional, boolean, default `false`): If the two images have different sizes, scale "after" to the size of "before" instead of stopping. - `highlight` (optional, string, default `"#ff0000"`): Colour that marks the differing pixels. #### Example 1: An image against itself Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"before":"DROP_1/f1","after":"DROP_2/f1"},"wait":30} ``` #### Example 2: Different sizes, ignoring small changes Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"before":"DROP_1/f1","after":"DROP_2/f1","resize":true,"tolerance":16},"wait":30} ``` **What does the difference image show?** A faded grey copy of "after" with every differing pixel painted in the highlight colour. **Why do two JPEGs that look the same differ?** JPEG compression changes many pixels slightly. Set tolerance to about 10-20 to ignore changes too small to see. **What is the box?** The smallest rectangle that holds every differing pixel, as x, y, width and height from the top left corner. **Is transparency compared?** Yes. A pixel whose transparency changed counts as different. ### Image info Show the facts about an image: dimensions, format, colour mode, transparency, animation frames, DPI and metadata. - Endpoint: `POST https://setget.net/v1/tools/image-info` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-info` - Page: https://setget.net/tools/image-info #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to inspect. #### Example 1: Inspect a PNG Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Inspect an animated GIF Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **What is the difference between size and dimensions?** Dimensions are the width and height in pixels. Size is how many bytes the file takes on disk. **What does "uses transparency" mean?** Some images can store transparency but every pixel is solid. "transparency" says the image can store it; "uses_transparency" says some pixels really are see-through. **Why is there no DPI?** Many images do not record one. DPI only matters for printing; screens ignore it. ### Image to ASCII art Convert an image to ASCII art text at a chosen width, using simple, detailed or block characters. - Endpoint: `POST https://setget.net/v1/tools/image-ascii-art` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-ascii-art` - Page: https://setget.net/tools/image-ascii-art #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to draw with characters. - `width` (optional, integer, default `80`, range 10..300): Width of the art in characters. - `charset` (optional, string, one of `simple`, `detailed`, `blocks`, default `"simple"`): simple uses 10 characters, detailed uses 70 for smoother shading, blocks uses shaded block characters. - `invert` (optional, boolean, default `false`): Swap light and dark. Turn on for light text on a dark background, such as a terminal. #### Example 1: 60 characters wide Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","width":60},"wait":30} ``` #### Example 2: Block characters for a dark terminal Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","width":40,"charset":"blocks","invert":true},"wait":30} ``` **Why does it look stretched or squashed?** Characters are about twice as tall as they are wide, so the art uses half as many lines as a square grid would. View it in a monospaced font. **When do I need invert?** By default dark parts of the image get the densest characters, which suits dark text on a light page. Turn invert on for light text on a dark background. **What happens to transparency?** Transparent areas are treated as white. ### Image to data URI Turn an image into a base64 data URI, with HTML img and CSS background-image snippets ready to paste. - Endpoint: `POST https://setget.net/v1/tools/image-to-data-uri` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-to-data-uri` - Page: https://setget.net/tools/image-to-data-uri #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to encode, up to 512 KB. SVG files work too. - `alt` (optional, string, default `""`): Alternative text for the HTML img tag. #### Example 1: Inline a small PNG Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Inline an SVG Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.svg](https://setget.net/static/samples/sample.svg). ```json {"input":{"image":"DROP_ID/f1","alt":"Logo"},"wait":30} ``` **Why is there a 512 KB limit?** A data URI is a third larger than the file and cannot be cached on its own, so large ones make pages slower. Compress or resize the image first, or link to it as a normal file. **Where are the snippets?** In the result, and also as three result files: the data URI as text, the HTML and the CSS. For images over 200 KB the HTML and CSS are in the files only, to keep the result small. **When should I inline an image?** For small icons and placeholders, where saving a network request matters more than the extra size. ### Invert image colours Invert the colours of an image to make a negative: black becomes white, blue becomes yellow. Transparency is kept. - Endpoint: `POST https://setget.net/v1/tools/image-invert` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-invert` - Page: https://setget.net/tools/image-invert #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to invert. #### Example 1: Make a negative Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Invert and keep transparency Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **What happens to transparent areas?** They stay exactly as transparent as they were. Only the colours are inverted. **Can I get the original back?** Inverting twice gives the original for exact formats such as PNG. A JPG changes very slightly each time it is saved. ### Make a GIF from images Build an animated GIF, WebP or PNG from 2 to 100 images played in order, with a chosen delay and loop count. - Endpoint: `POST https://setget.net/v1/tools/gif-from-images` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `gif-from-images` - Page: https://setget.net/tools/gif-from-images #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `images` (required, files, a list of file references such as `["DROP_ID/f1"]`, at most 100, accepts image): The frames, 2 to 100 images, in the order they should play. - `delay_ms` (optional, integer, default `500`, range 20..10000): Time each frame is shown, in milliseconds. - `loop` (optional, integer, default `0`, range 0..1000): How many times the animation repeats. 0 repeats forever. - `format` (optional, string, one of `gif`, `webp`, `png`, default `"gif"`): File format of the animation. png makes an animated PNG. - `width` (optional, integer, default `0`, range 0..4000): Width of the animation in pixels. 0 uses the width of the first image. Every frame is scaled to fit and centred. - `background` (optional, string, default `"#ffffff"`): Colour behind frames that have a different shape from the first. A hex colour, a colour name or transparent. #### Example 1: Three-frame GIF Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg), `DROP_3/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1","DROP_3/f1"]},"wait":30} ``` #### Example 2: Animated WebP, 200 pixels wide Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1"],"delay_ms":250,"format":"webp","width":200},"wait":30} ``` **What size is the animation?** The first image sets the shape. Every other image is scaled to fit inside it and centred, with the background colour filling any gap. Nothing is cropped. **Which format should I pick?** GIF plays everywhere but is limited to 256 colours per frame. WebP and animated PNG keep full colour and are usually smaller. **Why does my GIF have fewer frames than images?** Two identical images in a row are stored as one frame shown for twice as long. It plays the same. ### Make a thumbnail Make a small square thumbnail or preview of an image at a chosen size, cropped to fill or fitted inside a box. - Endpoint: `POST https://setget.net/v1/tools/image-thumbnail` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-thumbnail` - Page: https://setget.net/tools/image-thumbnail #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to make a thumbnail of. - `size` (optional, integer, default `256`, range 8..2048): Side of the thumbnail in pixels. - `fit` (optional, string, one of `cover`, `contain`, default `"cover"`): cover crops the middle to an exact square. contain keeps the whole image and its proportions, fitted inside a square of this size. - `format` (optional, string, one of `png`, `jpg`, `webp`, `avif`, `gif`, `bmp`, `tiff`, `ico`): Format of the result. Leave empty to keep the format of the upload. #### Example 1: 128 pixel square thumbnail Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","size":128},"wait":30} ``` #### Example 2: Whole image inside 128 pixels, as WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","size":128,"fit":"contain","format":"webp"},"wait":30} ``` **Is a small image enlarged?** With cover, yes: the result is always exactly the size you ask for. With contain, no: an image already smaller than the box is left at its own size. **How do I make a round avatar?** Make a square thumbnail here, then pass it to the round corners tool with circle set to true. ### Optimize an animated GIF Shrink an animated GIF by using fewer colours, a smaller size or fewer frames, and see how much was saved. - Endpoint: `POST https://setget.net/v1/tools/gif-optimize` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `gif-optimize` - Page: https://setget.net/tools/gif-optimize #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The GIF to shrink. - `colors` (optional, integer, default `128`, range 2..256): Colours in the result, 2-256. Fewer colours make a smaller file. - `scale` (optional, integer, default `100`, range 5..100): Size of the result as a percentage of the original width and height. - `drop_frames` (optional, integer, default `1`, range 1..20): Keep every Nth frame: 1 keeps all, 2 keeps every second frame. Kept frames are shown longer, so the animation runs at the same speed. #### Example 1: Reduce to 32 colours Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","colors":32},"wait":30} ``` #### Example 2: Half size, every second frame Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","colors":64,"scale":50,"drop_frames":2},"wait":30} ``` **Which setting saves the most?** Scale: 50 percent leaves a quarter of the pixels. Dropping frames comes next, then colours. For the smallest file of all, convert to WebP with animated-convert. **Can the result be larger than the original?** Occasionally, if the original was already tightly optimised and you change nothing. The result says so: saved_percent is then negative. **Does dropping frames speed the GIF up?** No. Each kept frame is shown for the time of the frames it replaces. **Does it work on WebP animations?** No, GIF only. Use animated-convert for other formats. ### Perceptual image hash Compute aHash, dHash and pHash fingerprints of an image, and compare two images to see if they look alike. - Endpoint: `POST https://setget.net/v1/tools/image-perceptual-hash` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-perceptual-hash` - Page: https://setget.net/tools/image-perceptual-hash #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to fingerprint. - `compare` (optional, file, a file reference such as `DROP_ID/f1`, accepts image): A second image to compare with the first. #### Example 1: Fingerprints of one image Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Do two images look alike? Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_1/f1","compare":"DROP_2/f1"},"wait":30} ``` **How is this different from a checksum?** A checksum changes completely if one byte changes. A perceptual hash stays nearly the same when an image is resized, recompressed or saved in another format. **How do I read the distance?** It is the number of bits, out of 64, that differ between two hashes. 0-2 means the images look the same, up to about 12 means similar, and more means different. **Which hash should I use?** pHash is the most robust. dHash is good at spotting small edits. aHash is the simplest and the most forgiving. The verdict uses pHash and dHash together. **Does it detect rotated or cropped copies?** No. Rotation, mirroring and heavy cropping change the hashes. ### Read EXIF data Read the EXIF data of a photo: camera, lens, date taken, exposure settings and GPS location as decimal degrees. - Endpoint: `POST https://setget.net/v1/tools/image-read-exif` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-read-exif` - Page: https://setget.net/tools/image-read-exif #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The photo to read. #### Example: Read a photo Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **Why is the result empty?** The image has no EXIF data. Screenshots, drawings and most images saved from social media or chat apps have had it removed. **How do I use the GPS numbers?** Latitude and longitude are decimal degrees; paste them as "latitude, longitude" into any map. Negative latitude is south and negative longitude is west. **Is the date in my time zone?** It is the camera clock at the moment of the photo. A time zone is added only when the camera recorded one. ### Redact part of an image Hide part of an image, such as a face, name, address or number, with a solid box, a blur or pixelation. - Endpoint: `POST https://setget.net/v1/tools/image-redact` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-redact` - Page: https://setget.net/tools/image-redact #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to redact. - `x` (required, integer, range 0..): Left edge of the area to hide, in pixels from the left of the image. - `y` (required, integer, range 0..): Top edge of the area to hide, in pixels from the top of the image. - `width` (required, integer, range 1..): Width of the area in pixels. - `height` (required, integer, range 1..): Height of the area in pixels. - `mode` (optional, string, one of `box`, `blur`, `pixelate`, default `"box"`): box covers the area with a solid colour and cannot be undone. blur and pixelate look softer but can sometimes be reversed. - `strength` (optional, integer, default `20`, range 2..200): For blur, how far each pixel is smeared; for pixelate, the size of the blocks. In pixels. - `color` (optional, string, default `"#000000"`): Colour of the box. A hex colour such as #ff8800, or a name such as white or black. #### Example 1: Black box Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","x":60,"y":40,"width":120,"height":60},"wait":30} ``` #### Example 2: Pixelate an area Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","x":60,"y":40,"width":120,"height":60,"mode":"pixelate","strength":12},"wait":30} ``` #### Example 3: Blur an area Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","x":0,"y":0,"width":100,"height":100,"mode":"blur"},"wait":30} ``` **Which mode is safe for secrets?** box. A solid box replaces the pixels completely, so nothing is left to recover. Use it for passwords, account numbers, addresses and anything else that must stay hidden. **Can blur or pixelation be reversed?** Sometimes, yes. Blurred or pixelated text and numbers have been recovered by software that tries likely values and compares the result. Use blur and pixelate for looks, such as softening a background face, not for secrets. **Is the hidden part still in the file somewhere?** No. The pixels are changed for good, and the camera data and embedded preview pictures of the original are removed from the result. **What if the area runs past the edge?** It is cut off at the edge of the image. The result reports the area that was covered. ### Remove a solid background Make a plain, single-colour background transparent, for logos, icons, scans and product shots. - Endpoint: `POST https://setget.net/v1/tools/image-remove-background` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-remove-background` - Page: https://setget.net/tools/image-remove-background #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image with a plain background. - `color` (optional, string, default `""`): The background colour to remove, as a hex colour or colour name. Leave empty to take it from the corners of the image. - `tolerance` (optional, integer, default `10`, range 0..100): How close a colour must be to count as background: 0 is an exact match, 100 matches everything. - `feather` (optional, integer, default `0`, range 0..20): Soften the cut edge over this many pixels. 0 leaves a hard edge. - `contiguous` (optional, boolean, default `true`): Only remove background connected to the edge of the image, so the same colour inside the subject is kept. Turn off to remove the colour everywhere. - `format` (optional, string, one of `png`, `webp`, default `"png"`): File format of the result. Both keep transparency. #### Example 1: Remove the colour found in the corners Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Remove white everywhere, with a soft edge Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","color":"#ffffff","tolerance":20,"feather":1,"contiguous":false},"wait":30} ``` **Does this work on photos?** Only if the background is one flat colour, such as a white studio backdrop or a green screen. It matches a colour; it does not recognise people or objects. Busy or graduated backgrounds need a dedicated cut-out program. **Parts of my subject became transparent. Why?** They are close to the background colour and touch the edge of it. Lower the tolerance, and keep contiguous on so matching colours inside the subject survive. **There is a thin halo around the subject. What helps?** Raise the tolerance a little and set feather to 1 or 2 to soften the edge. **How is the background colour chosen?** From the four corner pixels: the colour most of them share. Give "color" yourself if the corners are not background. ### Remove image metadata Remove EXIF data, GPS location, camera details, comments and colour profiles from a photo before sharing it. - Endpoint: `POST https://setget.net/v1/tools/image-strip-metadata` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-strip-metadata` - Page: https://setget.net/tools/image-strip-metadata #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to clean. - `keep_color_profile` (optional, boolean, default `false`): Keep the colour profile so colours look the same on wide-colour screens. It says nothing about you or where the photo was taken. #### Example: Clean a photo Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **What is removed?** Camera make and model, date, GPS position, embedded preview pictures, editing-software notes, comments and the colour profile. The result lists what was found. **Will my photo turn sideways?** No. If the camera stored the photo sideways with a note to rotate it, the pixels are turned upright first, so it still shows the right way up. **Does it change the picture?** PNG, GIF and other exact formats keep every pixel. A JPG is saved again at quality 95, which changes it very slightly. **Does this hide what is in the photo?** No. It removes hidden data only. Use the redact tool to cover faces, names or numbers. ### Remove transparency Replace the transparent areas of a PNG, WebP or GIF with a solid background colour such as white. - Endpoint: `POST https://setget.net/v1/tools/image-flatten` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-flatten` - Page: https://setget.net/tools/image-flatten #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image with transparency. - `background` (optional, string, default `"#ffffff"`): Solid colour to put behind the image. A hex colour such as #ff8800, or a name such as white or black. #### Example 1: Onto white Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` #### Example 2: Onto a dark colour Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"image":"DROP_ID/f1","background":"#1e293b"},"wait":30} ``` **Why would I remove transparency?** Some apps and printers show transparent areas as black, or reject such files. A solid background makes the image look the same everywhere. **What about half-transparent pixels?** They are blended with the background colour, so soft edges and shadows stay smooth. **What if my image has no transparency?** It comes back unchanged in appearance, and the result says no transparency was found. ### Resize image Resize an image to a width, a height or both, keeping its proportions, cropping to fill, or stretching. - Endpoint: `POST https://setget.net/v1/tools/image-resize` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-resize` - Page: https://setget.net/tools/image-resize #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to resize. - `width` (optional, integer, range 1..16000): New width in pixels. Leave empty to work it out from the height. - `height` (optional, integer, range 1..16000): New height in pixels. Leave empty to work it out from the width. - `fit` (optional, string, one of `contain`, `cover`, `stretch`, default `"contain"`): Used when both width and height are given. contain fits inside the box and keeps the proportions, cover fills the box and crops the overflow, stretch forces the exact size. - `upscale` (optional, boolean, default `false`): Allow the image to be made larger than it is. Off by default, because enlarging makes images blurry. - `format` (optional, string, one of `png`, `jpg`, `webp`, `avif`, `gif`, `bmp`, `tiff`, `ico`): Format of the result. Leave empty to keep the format of the upload. #### Example 1: Scale to 320 pixels wide Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","width":320},"wait":30} ``` #### Example 2: Fill a 200 x 200 square Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","width":200,"height":200,"fit":"cover"},"wait":30} ``` #### Example 3: Shrink an animated GIF Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1","width":80},"wait":30} ``` **What is the difference between contain, cover and stretch?** With a 200 x 200 box and a 640 x 400 photo: contain gives 200 x 125 with nothing cut off, cover gives exactly 200 x 200 with the sides cropped, and stretch gives 200 x 200 with the picture squashed. **Why is my result smaller than I asked for?** The image was smaller than the requested size and upscale is off. Set upscale to true to enlarge it. **Are animated GIFs resized?** Yes. Every frame is resized and the timing is kept. ### Rotate and flip image Rotate an image by any angle and flip it horizontally or vertically. Quarter turns keep every pixel exact. - Endpoint: `POST https://setget.net/v1/tools/image-rotate-flip` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-rotate-flip` - Page: https://setget.net/tools/image-rotate-flip #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to rotate or flip. - `angle` (optional, number, default `0`, range -3600..3600): Degrees to turn the image clockwise. Use a negative number to turn it anticlockwise. - `flip` (optional, string, one of `none`, `horizontal`, `vertical`, default `"none"`): Mirror the image after rotating: horizontal swaps left and right, vertical swaps top and bottom. - `background` (optional, string, default `"transparent"`): Colour for the empty corners left by a turn that is not a quarter turn. transparent becomes white in formats without transparency, such as JPG. A hex colour such as #ff8800, or a name such as white or black. - `expand` (optional, boolean, default `true`): Grow the canvas so the whole turned image fits. Turn off to keep the original size and cut off the corners. Quarter turns always keep the whole image. #### Example 1: Quarter turn clockwise Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","angle":90},"wait":30} ``` #### Example 2: Mirror left to right Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","flip":"horizontal"},"wait":30} ``` #### Example 3: Tilt 15 degrees on white Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","angle":15,"background":"#ffffff"},"wait":30} ``` **Which way is positive?** Clockwise. An angle of 90 turns the top of the image to the right; -90 turns it to the left. **Why did my image get larger after rotating?** A tilted rectangle needs a bigger canvas to fit. Set expand to false to keep the original size instead; the corners are then cut off. **My photo shows sideways in some apps. Will this fix it?** Yes. The photo is first turned the way the camera recorded it should be shown, then your rotation is applied, and the result displays the same everywhere. ### Round image corners Round the corners of an image, or cut it into a circle for an avatar, with transparent corners as PNG or WebP. - Endpoint: `POST https://setget.net/v1/tools/image-round-corners` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-round-corners` - Page: https://setget.net/tools/image-round-corners #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to round. - `radius` (optional, string, default `"10%"`): How round the corners are: pixels such as 24, or a percentage of the shorter side such as 10%. 50% is the most a corner can be rounded. - `circle` (optional, boolean, default `false`): Cut the image into a circle: the middle square is kept and fully rounded. The radius is ignored. - `format` (optional, string, one of `png`, `webp`, default `"png"`): Format of the result. Both keep the corners transparent. #### Example 1: Softly rounded corners Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","radius":"10%"},"wait":30} ``` #### Example 2: Circle for an avatar Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","circle":true},"wait":30} ``` #### Example 3: 24 pixel corners as WebP Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","radius":"24","format":"webp"},"wait":30} ``` **Why can I not get a JPG back?** The corners have to be transparent and JPG cannot store transparency. To put the rounded image on a solid colour, pass the result to the flatten tool. **How is the circle cut?** The largest square from the middle of the image is kept, then rounded fully. Crop the image first if the subject is off-centre. ### Split a GIF into frames Extract every frame of an animated GIF, WebP or PNG as its own PNG file, with each frame's duration. - Endpoint: `POST https://setget.net/v1/tools/gif-split-frames` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `gif-split-frames` - Page: https://setget.net/tools/gif-split-frames #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The animation to split: GIF, animated WebP, animated PNG or AVIF. Up to 100 frames. #### Example: Eight frames of a GIF Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.gif](https://setget.net/static/samples/sample.gif). ```json {"input":{"image":"DROP_ID/f1"},"wait":30} ``` **What is the frame limit?** A job returns at most 100 files, so animations with more than 100 frames are refused. Thin one out first with gif-optimize and its drop_frames setting. **Are frames complete pictures?** Yes. GIFs often store only the part that changed; each PNG here is the full picture as it appears at that moment. **How are the files named?** After the animation, with the frame number: name-001.png, name-002.png and so on. ### Split an image into tiles Cut an image into a grid of tiles by rows and columns, or into tiles of a fixed pixel size. - Endpoint: `POST https://setget.net/v1/tools/image-tile-split` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `image-tile-split` - Page: https://setget.net/tools/image-tile-split #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `image` (required, file, a file reference such as `DROP_ID/f1`, accepts image): The image to cut up. - `rows` (optional, integer, default `2`, range 1..100): Number of rows in the grid. - `columns` (optional, integer, default `2`, range 1..100): Number of columns in the grid. - `tile_width` (optional, integer, default `0`, range 0..16000): Width of each tile in pixels. Set this or tile_height to cut by size instead of by rows and columns. 0 means the full width. - `tile_height` (optional, integer, default `0`, range 0..16000): Height of each tile in pixels. 0 means the full height. #### Example 1: 2 rows of 3 tiles Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"image":"DROP_ID/f1","rows":2,"columns":3},"wait":30} ``` #### Example 2: 256 pixel tiles Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"image":"DROP_ID/f1","tile_width":256,"tile_height":256},"wait":30} ``` **What is the limit?** At most 100 tiles per run. Ask for fewer rows and columns, or larger tiles. **What if the image does not divide evenly?** By rows and columns, tiles differ by at most one pixel. By pixel size, the last tile in each row and column is smaller. **How are the tiles named?** After the image, with row and column counted from the top left: name-r1-c1, name-r1-c2 and so on. **Which format are the tiles?** PNG, JPEG and WebP images keep their format. Everything else becomes PNG. ### Sprite sheet maker Pack up to 64 images into one sprite sheet, with the position of every image and ready-to-paste CSS. - Endpoint: `POST https://setget.net/v1/tools/sprite-sheet` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `sprite-sheet` - Page: https://setget.net/tools/sprite-sheet #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `images` (required, files, a list of file references such as `["DROP_ID/f1"]`, at most 64, accepts image): The images to pack, up to 64. - `layout` (optional, string, one of `shelf`, `grid`, default `"shelf"`): shelf packs images tightly in rows, tallest first. grid gives every image a cell of the same size, in upload order. - `padding` (optional, integer, default `2`, range 0..64): Empty pixels between images and around the edge. - `max_width` (optional, integer, default `0`, range 0..16000): Widest the sheet may be in pixels, for the shelf layout. 0 picks a roughly square sheet. - `class_prefix` (optional, string, default `"sprite"`): Start of the CSS class names, for example sprite gives .sprite-NAME. - `format` (optional, string, one of `png`, `webp`, default `"png"`): File format of the sheet. #### Example 1: Pack three images Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png), `DROP_3/f1` = [sample.jpg](https://setget.net/static/samples/sample.jpg). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1","DROP_3/f1"]},"wait":30} ``` #### Example 2: Even grid without padding Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.png](https://setget.net/static/samples/sample.png), `DROP_2/f1` = [sample-transparent.png](https://setget.net/static/samples/sample-transparent.png). ```json {"input":{"images":["DROP_1/f1","DROP_2/f1"],"layout":"grid","padding":0},"wait":30} ``` **How do I use the CSS?** Save the sheet next to your stylesheet, paste the CSS, then give an element the base class and one image class, for example class="sprite sprite-logo". **What are the frame names?** The file name of each image without its extension, in lower case with other characters turned into dashes. Two files with the same name get -2, -3 and so on. **Does the padding matter?** A pixel or two of padding stops neighbouring images from bleeding into each other when the sheet is scaled. ### SVG optimizer Shrink an SVG file by removing comments, editor data, metadata and spare white space, and rounding long decimals. Removes what a browser never draws: comments, the XML declaration and doctype, metadata blocks, Inkscape, Sodipodi, Sketch, Figma and Illustrator editor data, empty groups, and white space between elements. Long decimal numbers in coordinates are rounded. Text, styles, ids, transforms and the viewBox are left exactly as they are, so the drawing looks the same. - Endpoint: `POST https://setget.net/v1/tools/svg-optimize` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `svg-optimize` - Page: https://setget.net/tools/svg-optimize #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `svg` (required, file, a file reference such as `DROP_ID/f1`, accepts image, text): The SVG file to shrink. - `precision` (optional, integer, default `3`, range 1..8): Decimal places kept in coordinates and sizes. Numbers below 1 keep this many significant digits instead. - `strip_scripts` (optional, boolean, default `false`): Also remove scripts, event handlers such as onclick, and javascript: links. #### Example 1: Clean up an SVG Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.svg](https://setget.net/static/samples/sample.svg). ```json {"input":{"svg":"DROP_ID/f1"},"wait":30} ``` #### Example 2: One decimal place, scripts removed Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.svg](https://setget.net/static/samples/sample.svg). ```json {"input":{"svg":"DROP_ID/f1","precision":1,"strip_scripts":true},"wait":30} ``` **Will my drawing look different?** It should not. Only things with no effect on the picture are removed. Rounding moves points by less than a thousandth of a unit at the default precision; raise "precision" for drawings with a very small viewBox, or if you see any change. **Can I still edit the file in Inkscape or Illustrator?** Yes, it stays a normal SVG. Editor-only extras such as guides, grid settings and layer names are removed, so keep your original for editing. **What happens to scripts?** They are kept unless you set strip_scripts. The result always tells you how many scripts and event handlers the file holds, which is worth checking for files from other people. **Why was my file refused for "entities"?** Files that define their own entities () can be built to overload the program reading them, so they are not processed. Illustrator writes these when "Preserve Illustrator Editing Capabilities" is ticked: export again without it. **What is not done?** Ids are not renamed, shapes are not merged and styles are not rewritten. Those save more but can change how a drawing behaves on a page. ### SVG to PNG Render an SVG drawing to a PNG image at any size, with a transparent or solid background. - Endpoint: `POST https://setget.net/v1/tools/svg-to-png` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `svg-to-png` - Page: https://setget.net/tools/svg-to-png #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `svg` (required, file, a file reference such as `DROP_ID/f1`, accepts image, text): The SVG file to render. - `width` (optional, integer, range 1..16000): Width of the PNG in pixels. Leave empty to use the drawing's own size. - `height` (optional, integer, range 1..16000): Height of the PNG in pixels. With only one of width and height, the other follows the drawing's proportions. - `scale` (optional, number, default `1`, range 0.1..20): Multiply the drawing's own size, for example 2 for a sharper image. Ignored when width or height is given. - `background` (optional, string): Background colour such as #ffffff. Leave empty to keep it transparent. #### Example 1: Render 480 pixels wide Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.svg](https://setget.net/static/samples/sample.svg). ```json {"input":{"svg":"DROP_ID/f1","width":480},"wait":30} ``` #### Example 2: Double size on white Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.svg](https://setget.net/static/samples/sample.svg). ```json {"input":{"svg":"DROP_ID/f1","scale":2,"background":"#ffffff"},"wait":30} ``` **Does it load images or fonts the SVG links to?** No. Only what is embedded in the file is drawn. Linked images are left out and text uses a standard font. **How do I get a JPG or WebP instead?** Render to PNG here, then pass the result to image-convert. **What are the limits?** A 5 MB SVG file, and a result of at most 16,000 pixels per side and 50 megapixels. ## Data Convert, query, compare and validate JSON, YAML, TOML, XML and CSV. ### .env to JSON Parse a .env file into a JSON object, handling quotes, comments and "export" prefixes. - Endpoint: `POST https://setget.net/v1/tools/env-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `env-to-json` - Page: https://setget.net/tools/env-to-json #### Inputs - `text` (required, string): The contents of a .env file. Do not paste real secrets. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"# app\nPORT=8080\nexport NAME=\"My App\"\nDEBUG=true # on","indent":0}} ``` Result: ```json {"text":"{\"PORT\":\"8080\",\"NAME\":\"My App\",\"DEBUG\":\"true\"}","variables":3,"skipped":[]} ``` **Is it safe to paste secrets?** No. Input is processed in memory and not stored, but you should never send live credentials to any online tool. Replace them with placeholders first. ### CSV profiler Summarise each CSV column: type, blanks, distinct values, min, max, mean and top values. - Endpoint: `POST https://setget.net/v1/tools/csv-profile` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `csv-profile` - Page: https://setget.net/tools/csv-profile #### Inputs - `text` (required, string): The CSV to profile. - `delimiter` (optional, string, one of `auto`, `comma`, `tab`, `semicolon`, `pipe`, default `"auto"`): Column separator. "auto" detects it from the first line. #### Example ```json {"input":{"text":"city,temp\nOslo,4\nRome,18\nOslo,"}} ``` Result: ```json {"rows":3,"columns":[{"name":"city","type":"text","blank":0,"distinct":2,"top":[["Oslo",2],["Rome",1]]},{"name":"temp","type":"number","blank":1,"distinct":2,"min":4,"max":18,"mean":11.0,"top":[["4",1],["18",1]]}]} ``` ### CSV sort, filter and dedupe Sort CSV by a column, keep rows containing some text, pick columns and remove duplicate rows. - Endpoint: `POST https://setget.net/v1/tools/csv-sort-filter-dedupe` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `csv-sort-filter-dedupe` - Page: https://setget.net/tools/csv-sort-filter-dedupe #### Inputs - `text` (required, string): The CSV to process. - `delimiter` (optional, string, one of `auto`, `comma`, `tab`, `semicolon`, `pipe`, default `"auto"`): Column separator. "auto" detects it from the first line. - `sort_by` (optional, string): Column name to sort by. Numbers sort numerically. - `descending` (optional, boolean, default `false`): Sort from largest to smallest. - `contains` (optional, string): Keep only rows where some cell contains this text (case-insensitive). - `columns` (optional, array): Column names to keep, in this order. All columns when omitted. - `dedupe` (optional, boolean, default `false`): Remove rows that are exact duplicates of an earlier row. #### Example ```json {"input":{"text":"name,age\nSam,31\nAlex,27\nSam,31","sort_by":"age","dedupe":true}} ``` Result: ```json {"text":"name,age\nAlex,27\nSam,31\n","rows":2,"removed":1} ``` ### CSV to JSON Turn CSV rows into a JSON array of objects, detecting numbers, booleans and blanks. - Endpoint: `POST https://setget.net/v1/tools/csv-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `csv-to-json` - Page: https://setget.net/tools/csv-to-json #### Inputs - `text` (required, string): The CSV to convert. - `delimiter` (optional, string, one of `auto`, `comma`, `tab`, `semicolon`, `pipe`, default `"auto"`): Column separator. "auto" detects it from the first line. - `header` (optional, boolean, default `true`): The first row holds column names. - `infer_types` (optional, boolean, default `true`): Turn numbers, true/false and blanks into typed values. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"name,age\nSam,31\nAlex,27","indent":0}} ``` Result: ```json {"text":"[{\"name\":\"Sam\",\"age\":31},{\"name\":\"Alex\",\"age\":27}]","rows":2,"columns":["name","age"]} ``` **Why did "007" stay as text?** Values with leading zeros are usually codes, not numbers, so they are left alone. ### CSV to Markdown table Turn CSV into a Markdown table ready to paste into a README, issue or chat. - Endpoint: `POST https://setget.net/v1/tools/csv-to-markdown-table` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `csv-to-markdown-table` - Page: https://setget.net/tools/csv-to-markdown-table #### Inputs - `text` (required, string): The CSV to convert. - `delimiter` (optional, string, one of `auto`, `comma`, `tab`, `semicolon`, `pipe`, default `"auto"`): Column separator. "auto" detects it from the first line. - `align` (optional, string, one of `none`, `left`, `center`, `right`, default `"none"`): Column alignment. #### Example ```json {"input":{"text":"name,age\nSam,31"}} ``` Result: ```json {"text":"| name | age |\n| --- | --- |\n| Sam | 31 |"} ``` ### CSV transpose Swap the rows and columns of a CSV table. - Endpoint: `POST https://setget.net/v1/tools/csv-transpose` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `csv-transpose` - Page: https://setget.net/tools/csv-transpose #### Inputs - `text` (required, string): The CSV to transpose. - `delimiter` (optional, string, one of `auto`, `comma`, `tab`, `semicolon`, `pipe`, default `"auto"`): Column separator. "auto" detects it from the first line. #### Example ```json {"input":{"text":"a,b\n1,2\n3,4"}} ``` Result: ```json {"text":"a,1,3\nb,2,4\n"} ``` ### JSON diff Compare two JSON documents and get the differences as a JSON Patch (RFC 6902). - Endpoint: `POST https://setget.net/v1/tools/json-diff` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-diff` - Page: https://setget.net/tools/json-diff #### Inputs - `before` (required, string): The original JSON. - `after` (required, string): The changed JSON. #### Example ```json {"input":{"before":"{\"a\":1,\"b\":[1,2],\"c\":true}","after":"{\"a\":2,\"b\":[1,2,3]}"}} ``` Result: ```json {"equal":false,"patch":[{"op":"remove","path":"/c"},{"op":"replace","path":"/a","value":2},{"op":"add","path":"/b/-","value":3}],"summary":{"add":1,"remove":1,"replace":1}} ``` **Does key order matter?** No. Objects are compared by key, so reordering keys is not a difference. **How are arrays compared?** Position by position. An item inserted at the start shows as changes to every later position. ### JSON flatten Flatten nested JSON into one level of dotted keys such as "user.address.city". - Endpoint: `POST https://setget.net/v1/tools/json-flatten` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-flatten` - Page: https://setget.net/tools/json-flatten #### Inputs - `text` (required, string): The JSON to flatten. - `separator` (optional, string, default `"."`): Text placed between key parts. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"{\"user\":{\"name\":\"Sam\",\"tags\":[\"a\",\"b\"]}}","indent":0}} ``` Result: ```json {"text":"{\"user.name\":\"Sam\",\"user.tags.0\":\"a\",\"user.tags.1\":\"b\"}","keys":3} ``` ### JSON formatter Pretty-print JSON with your choice of indentation, optionally sorting keys. - Endpoint: `POST https://setget.net/v1/tools/json-format` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-format` - Page: https://setget.net/tools/json-format #### Inputs - `text` (required, string): The JSON to format. - `indent` (optional, integer, default `2`, range 1..8): Spaces per indentation level. - `sort_keys` (optional, boolean, default `false`): Sort object keys alphabetically. #### Example 1 ```json {"input":{"text":"{\"b\":1,\"a\":[1,2]}"}} ``` Result: ```json {"text":"{\n \"b\": 1,\n \"a\": [\n 1,\n 2\n ]\n}"} ``` #### Example 2: Sorted keys, four spaces ```json {"input":{"text":"{\"b\":1,\"a\":2}","sort_keys":true,"indent":4}} ``` Result: ```json {"text":"{\n \"a\": 2,\n \"b\": 1\n}"} ``` **Does formatting change my data?** No. Only whitespace changes, and key order if you ask for sorting. **What happens if the JSON is broken?** You get the line and column of the first fault, so you can fix it. ### JSON Merge Patch Merge a JSON Merge Patch (RFC 7386) into a document: objects merge, null deletes a key. - Endpoint: `POST https://setget.net/v1/tools/json-merge-patch` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-merge-patch` - Page: https://setget.net/tools/json-merge-patch #### Inputs - `text` (required, string): The JSON document to change. - `patch` (required, string): The merge patch. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"{\"a\":1,\"b\":{\"c\":2,\"d\":3}}","patch":"{\"b\":{\"c\":null,\"e\":4}}","indent":0}} ``` Result: ```json {"text":"{\"a\":1,\"b\":{\"d\":3,\"e\":4}}"} ``` **How do I replace an array?** Arrays are always replaced whole. Merge Patch cannot change a single array item. ### JSON minifier Remove all optional whitespace from JSON to make it as small as possible. - Endpoint: `POST https://setget.net/v1/tools/json-minify` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-minify` - Page: https://setget.net/tools/json-minify #### Inputs - `text` (required, string): The JSON to minify. #### Example ```json {"input":{"text":"{ \"a\": 1,\n \"b\": [1, 2] }"}} ``` Result: ```json {"text":"{\"a\":1,\"b\":[1,2]}","bytes_before":25,"bytes_after":17} ``` ### JSON Patch apply Apply a JSON Patch (RFC 6902) to a document: add, remove, replace, move, copy and test. - Endpoint: `POST https://setget.net/v1/tools/json-patch-apply` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-patch-apply` - Page: https://setget.net/tools/json-patch-apply #### Inputs - `text` (required, string): The JSON document to change. - `patch` (required, string): A JSON array of patch operations. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"{\"a\":1,\"b\":[1,2]}","patch":"[{\"op\":\"replace\",\"path\":\"/a\",\"value\":2},{\"op\":\"add\",\"path\":\"/b/-\",\"value\":3}]","indent":0}} ``` Result: ```json {"text":"{\"a\":2,\"b\":[1,2,3]}","applied":2} ``` ### JSON Schema generator Generate a JSON Schema from example JSON. Arrays of objects are merged into one item shape. - Endpoint: `POST https://setget.net/v1/tools/json-schema-infer` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-schema-infer` - Page: https://setget.net/tools/json-schema-infer #### Inputs - `text` (required, string): Example JSON to describe. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"[{\"id\":1,\"name\":\"a\"},{\"id\":2}]","indent":0}} ``` Result: ```json {"text":"{\"$schema\":\"https://json-schema.org/draft/2020-12/schema\",\"type\":\"array\",\"items\":{\"type\":\"object\",\"properties\":{\"id\":{\"type\":\"integer\"},\"name\":{\"type\":\"string\"}},\"required\":[\"id\"]}}"} ``` **How is "required" decided?** A key is required only if it appears in every example object. ### JSON Schema validator Validate JSON against a JSON Schema and list every violation with its location. - Endpoint: `POST https://setget.net/v1/tools/json-schema-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-schema-validate` - Page: https://setget.net/tools/json-schema-validate #### Inputs - `text` (required, string): The JSON document to validate. - `schema` (required, string): The JSON Schema. #### Example 1 ```json {"input":{"text":"{\"name\":\"Sam\",\"age\":\"31\"}","schema":"{\"type\":\"object\",\"required\":[\"name\",\"age\"],\"properties\":{\"age\":{\"type\":\"integer\"}}}"}} ``` Result: ```json {"valid":false,"errors":[{"path":"$.age","message":"'31' is not of type 'integer'","rule":"type"}]} ``` #### Example 2: A valid document ```json {"input":{"text":"{\"name\":\"Sam\",\"age\":31}","schema":"{\"type\":\"object\",\"required\":[\"name\"]}"}} ``` Result: ```json {"valid":true,"errors":[]} ``` **Which schema drafts work?** The draft named in "$schema" is used. Without one, the latest draft (2020-12) applies. **Are remote $ref URLs fetched?** No. This tool never makes network requests, so every $ref must point inside the schema. ### JSON to CSV Turn a JSON array of objects into CSV. Nested objects become dotted column names. - Endpoint: `POST https://setget.net/v1/tools/json-to-csv` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-to-csv` - Page: https://setget.net/tools/json-to-csv #### Inputs - `text` (required, string): A JSON array of objects (or one object). - `delimiter` (optional, string, one of `comma`, `tab`, `semicolon`, `pipe`, default `"comma"`): Column separator for the output. #### Example ```json {"input":{"text":"[{\"name\":\"Sam\",\"address\":{\"city\":\"Oslo\"}},{\"name\":\"Alex\"}]"}} ``` Result: ```json {"text":"name,address.city\nSam,Oslo\nAlex,\n","rows":2,"columns":["name","address.city"]} ``` ### JSON to TOML Convert a JSON object into TOML with tables and arrays of tables. - Endpoint: `POST https://setget.net/v1/tools/json-to-toml` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-to-toml` - Page: https://setget.net/tools/json-to-toml #### Inputs - `text` (required, string): A JSON object to convert. #### Example ```json {"input":{"text":"{\"title\":\"x\",\"owner\":{\"name\":\"Sam\",\"admin\":true}}"}} ``` Result: ```json {"text":"title = \"x\"\n\n[owner]\nname = \"Sam\"\nadmin = true\n"} ``` **Why does a null value fail?** TOML has no null. Remove the key or give it a value before converting. ### JSON to XML Convert JSON into indented XML. "@name" keys become attributes and arrays repeat the tag. - Endpoint: `POST https://setget.net/v1/tools/json-to-xml` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-to-xml` - Page: https://setget.net/tools/json-to-xml #### Inputs - `text` (required, string): The JSON to convert. - `root` (optional, string, default `"root"`): Root tag name, used when the JSON is not an object with a single key. #### Example ```json {"input":{"text":"{\"user\":{\"@id\":\"7\",\"name\":\"Sam\",\"tag\":[\"a\",\"b\"]}}"}} ``` Result: ```json {"text":"\n Sam\n a\n b\n"} ``` ### JSON to YAML Convert JSON into clean block-style YAML, keeping key order. - Endpoint: `POST https://setget.net/v1/tools/json-to-yaml` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-to-yaml` - Page: https://setget.net/tools/json-to-yaml #### Inputs - `text` (required, string): The JSON to convert. #### Example ```json {"input":{"text":"{\"name\":\"app\",\"ports\":[80,443]}"}} ``` Result: ```json {"text":"name: app\nports:\n- 80\n- 443\n"} ``` ### JSON unflatten Rebuild nested JSON from flat dotted keys. Numeric key parts become array positions. - Endpoint: `POST https://setget.net/v1/tools/json-unflatten` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-unflatten` - Page: https://setget.net/tools/json-unflatten #### Inputs - `text` (required, string): A flat JSON object with dotted keys. - `separator` (optional, string, default `"."`): Text that separates key parts. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"{\"user.name\":\"Sam\",\"user.tags.0\":\"a\",\"user.tags.1\":\"b\"}","indent":0}} ``` Result: ```json {"text":"{\"user\":{\"name\":\"Sam\",\"tags\":[\"a\",\"b\"]}}"} ``` ### JSON validator Check whether text is valid JSON and get the exact line and column of the first error. - Endpoint: `POST https://setget.net/v1/tools/json-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `json-validate` - Page: https://setget.net/tools/json-validate #### Inputs - `text` (required, string): The text to check. #### Example 1 ```json {"input":{"text":"{\"a\": 1}"}} ``` Result: ```json {"valid":true,"type":"object","error":null} ``` #### Example 2: A misspelt literal ```json {"input":{"text":"{\"a\": tru}"}} ``` Result: ```json {"valid":false,"type":null,"error":{"message":"Expecting value","line":1,"column":7}} ``` ### JSONPath query Pull values out of JSON with a JSONPath expression such as $.items[*].name. - Endpoint: `POST https://setget.net/v1/tools/jsonpath-query` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `jsonpath-query` - Page: https://setget.net/tools/jsonpath-query #### Inputs - `text` (required, string): The JSON to search. - `path` (required, string): JSONPath expression starting with $. #### Example 1 ```json {"input":{"text":"{\"items\":[{\"name\":\"a\",\"n\":1},{\"name\":\"b\",\"n\":2}]}","path":"$.items[*].name"}} ``` Result: ```json {"matches":["a","b"],"count":2} ``` #### Example 2: Recursive descent ```json {"input":{"text":"{\"a\":{\"id\":1,\"b\":{\"id\":2}}}","path":"$..id"}} ``` Result: ```json {"matches":[1,2],"count":2} ``` **Which JSONPath features are supported?** Names, quoted names, indexes, slices, wildcards and recursive descent (..). Filter expressions such as ?(@.price > 10) are not supported. ### NDJSON converter Convert between a JSON array and newline-delimited JSON (NDJSON, JSON Lines). - Endpoint: `POST https://setget.net/v1/tools/ndjson-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ndjson-convert` - Page: https://setget.net/tools/ndjson-convert #### Inputs - `text` (required, string): A JSON array, or NDJSON with one value per line. - `to` (optional, string, one of `auto`, `ndjson`, `json`, default `"auto"`): Output format. #### Example 1 ```json {"input":{"text":"[{\"a\":1},{\"a\":2}]"}} ``` Result: ```json {"text":"{\"a\":1}\n{\"a\":2}\n","format":"ndjson","records":2} ``` #### Example 2 ```json {"input":{"text":"{\"a\":1}\n{\"a\":2}","to":"json"}} ``` Result: ```json {"text":"[{\"a\":1},{\"a\":2}]","format":"json","records":2} ``` ### Query string parser Decode a URL query string into JSON. Repeated names become arrays. - Endpoint: `POST https://setget.net/v1/tools/querystring-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `querystring-parse` - Page: https://setget.net/tools/querystring-parse #### Inputs - `text` (required, string): A query string or a full URL. #### Example ```json {"input":{"text":"https://example.com/s?q=hello+world&tag=a&tag=b"}} ``` Result: ```json {"params":{"q":"hello world","tag":["a","b"]},"count":3} ``` ### TOML to JSON Convert TOML into JSON. Dates and times become ISO strings. - Endpoint: `POST https://setget.net/v1/tools/toml-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `toml-to-json` - Page: https://setget.net/tools/toml-to-json #### Inputs - `text` (required, string): The TOML to convert. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"title = \"x\"\n[owner]\nname = \"Sam\"","indent":0}} ``` Result: ```json {"text":"{\"title\":\"x\",\"owner\":{\"name\":\"Sam\"}}"} ``` ### XML formatter Pretty-print XML with consistent indentation, or collapse it to a single line. - Endpoint: `POST https://setget.net/v1/tools/xml-format` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `xml-format` - Page: https://setget.net/tools/xml-format #### Inputs - `text` (required, string): The XML to format. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 removes whitespace between tags. #### Example ```json {"input":{"text":"1"}} ``` Result: ```json {"text":"\n 1\n \n"} ``` ### XML to JSON Convert XML into JSON. Attributes become "@name" keys and repeated tags become arrays. - Endpoint: `POST https://setget.net/v1/tools/xml-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `xml-to-json` - Page: https://setget.net/tools/xml-to-json #### Inputs - `text` (required, string): The XML to convert. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"Samab","indent":0}} ``` Result: ```json {"text":"{\"user\":{\"@id\":\"7\",\"name\":\"Sam\",\"tag\":[\"a\",\"b\"]}}"} ``` **How are attributes represented?** As keys starting with "@". Text next to child elements is stored under "#text". ### YAML to JSON Convert YAML into JSON. Dates become ISO strings; anchors and aliases are resolved. - Endpoint: `POST https://setget.net/v1/tools/yaml-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `yaml-to-json` - Page: https://setget.net/tools/yaml-to-json #### Inputs - `text` (required, string): The YAML to convert. - `indent` (optional, integer, default `2`, range 0..8): Spaces per level. 0 gives one compact line. #### Example ```json {"input":{"text":"name: app\nports:\n - 80\n - 443","indent":0}} ``` Result: ```json {"text":"{\"name\":\"app\",\"ports\":[80,443]}"} ``` ### YAML validator Check whether text is valid YAML and see where the first error is. - Endpoint: `POST https://setget.net/v1/tools/yaml-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `yaml-validate` - Page: https://setget.net/tools/yaml-validate #### Inputs - `text` (required, string): The YAML to check. #### Example ```json {"input":{"text":"a: 1\nb:\n - x"}} ``` Result: ```json {"valid":true,"type":"object","documents":1,"error":null} ``` ## Text Compare, clean, search, reshape and measure text. ### Apply a patch Apply a unified diff to a text and get the patched result, with clear errors when a hunk does not fit. - Endpoint: `POST https://setget.net/v1/tools/patch-apply` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `patch-apply` - Page: https://setget.net/tools/patch-apply #### Inputs - `text` (required, string): The original text to change. - `patch` (required, string): A unified diff for one file. #### Example 1 ```json {"input":{"text":"apple\nbanana\ncherry\n","patch":"--- before\n+++ after\n@@ -1,3 +1,4 @@\n apple\n-banana\n+blueberry\n cherry\n+date\n"}} ``` Result: ```json {"text":"apple\nblueberry\ncherry\ndate\n","hunks":1,"added":2,"removed":1} ``` #### Example 2: A patch without file headers ```json {"input":{"text":"one\ntwo\n","patch":"@@ -2 +2 @@\n-two\n+2\n"}} ``` Result: ```json {"text":"one\n2\n","hunks":1,"added":1,"removed":1} ``` **What if the line numbers in the patch are slightly off?** Each hunk is first tried at the line its header names. If the text there is different, the nearest place where the hunk fits exactly is used. **Why does a hunk not match?** The text is not the version the patch was made from, or the patch lost its leading spaces when it was copied. The error names the hunk and the first line that differs. **Can it apply a patch for several files?** No. Apply the part for each file to that file's text. ### Case converter Convert text between lower, UPPER, Title, Sentence, camelCase, PascalCase, snake_case, kebab-case and more. - Endpoint: `POST https://setget.net/v1/tools/case-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `case-convert` - Page: https://setget.net/tools/case-convert #### Inputs - `text` (required, string): The text or names to convert. - `to` (required, string, one of `lower`, `upper`, `title`, `sentence`, `camel`, `pascal`, `snake`, `kebab`, `constant`, `dot`): The case to convert to. #### Example 1 ```json {"input":{"text":"Hello world_fooBar","to":"camel"}} ``` Result: ```json {"text":"helloWorldFooBar"} ``` #### Example 2: Acronyms stay together ```json {"input":{"text":"HTTPServer error-count","to":"snake"}} ``` Result: ```json {"text":"http_server_error_count"} ``` #### Example 3: Sentence case ```json {"input":{"text":"the end. a new START","to":"sentence"}} ``` Result: ```json {"text":"The end. A new start"} ``` **How are words found for camel, snake and the other name styles?** Words are split at spaces, punctuation, underscores and hyphens, and where a lower-case letter is followed by a capital. Each line is converted on its own. **Why did sentence case lower a name such as "London"?** Sentence case lowers everything and then capitalises the first letter of each sentence. It cannot know which words are names. **Does title case skip small words like "of"?** No. Every word gets a capital letter. ### Escape and unescape text Escape or unescape text for HTML, URLs, JSON strings, shell commands, regular expressions and CSV fields. - Endpoint: `POST https://setget.net/v1/tools/escape-unescape` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `escape-unescape` - Page: https://setget.net/tools/escape-unescape #### Inputs - `text` (required, string): The text to escape or unescape. - `format` (required, string, one of `html`, `url`, `json`, `shell`, `regex`, `csv`): What the text is being prepared for. - `direction` (optional, string, one of `escape`, `unescape`, default `"escape"`): Escape plain text, or turn escaped text back into plain text. #### Example 1 ```json {"input":{"text":"Tom & Jerry","format":"html"}} ``` Result: ```json {"text":"<a href="x">Tom & Jerry</a>"} ``` #### Example 2: Decoding a URL ```json {"input":{"text":"caf%C3%A9%20au%20lait","format":"url","direction":"unescape"}} ``` Result: ```json {"text":"café au lait"} ``` #### Example 3: Quoting for a POSIX shell ```json {"input":{"text":"it's here","format":"shell"}} ``` Result: ```json {"text":"'it'\"'\"'s here'"} ``` #### Example 4: A JSON string literal ```json {"input":{"text":"Line 1\nSay \"hi\"","format":"json"}} ``` Result: ```json {"text":"\"Line 1\\nSay \\\"hi\\\"\""} ``` **Which shells does the shell format suit?** POSIX shells such as sh, bash and zsh. It is not correct for Windows cmd or PowerShell. **Does URL unescape turn "+" into a space?** No. "+" only means a space in form data. Here it stays a plus sign; "%20" becomes a space. **Does JSON escape include the surrounding quotes?** Yes, the result is a complete JSON string. Unescape accepts the text with or without them. **What does shell unescape return for several words?** The words are returned separately in "words", and joined by single spaces in "text". ### Extract emails and URLs Pull every email address and web link out of a text, and optionally IPv4 addresses and hashtags. - Endpoint: `POST https://setget.net/v1/tools/extract-emails-urls` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `extract-emails-urls` - Page: https://setget.net/tools/extract-emails-urls #### Inputs - `text` (required, string): The text to search. - `include_ips` (optional, boolean, default `false`): Also list IPv4 addresses. - `include_hashtags` (optional, boolean, default `false`): Also list hashtags such as #news. #### Example 1 ```json {"input":{"text":"Mail sam@example.com or alex@example.org, see https://example.com/docs."}} ``` Result: ```json {"emails":["sam@example.com","alex@example.org"],"urls":["https://example.com/docs"],"count":3} ``` #### Example 2: With IP addresses and hashtags ```json {"input":{"text":"Server 10.0.0.5 is down (www.example.com/status) #outage #outage","include_ips":true,"include_hashtags":true}} ``` Result: ```json {"emails":[],"urls":["www.example.com/status"],"ips":["10.0.0.5"],"hashtags":["#outage"],"count":3} ``` **Which links are found?** Links that start with http://, https://, ftp:// or www. A bare domain such as "example.com" is not listed. **Are duplicates removed?** Yes. Each value is listed once, in the order it first appears. **Does it check that an address really exists?** No. It only looks at the shape of the text and never contacts the network. ### Find and replace Replace every occurrence of a word or phrase in text, exactly as typed, with a count of the changes. - Endpoint: `POST https://setget.net/v1/tools/find-replace` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `find-replace` - Page: https://setget.net/tools/find-replace #### Inputs - `text` (required, string): The text to change. - `find` (required, string): The exact text to look for. No special characters. - `replace` (optional, string, default `""`): The text to put in its place. Leave empty to delete. - `case_sensitive` (optional, boolean, default `true`): Treat upper and lower case as different. - `whole_word` (optional, boolean, default `false`): Only match when the text is not part of a longer word. #### Example 1 ```json {"input":{"text":"Cat, concat, cat.","find":"cat","replace":"dog","case_sensitive":false,"whole_word":true}} ``` Result: ```json {"text":"dog, concat, dog.","count":2} ``` #### Example 2: Special characters are plain text ```json {"input":{"text":"price: $5.00 (was $7.00)","find":"$","replace":"USD "}} ``` Result: ```json {"text":"price: USD 5.00 (was USD 7.00)","count":2} ``` **Can I use wildcards or patterns?** Not here. Every character is matched exactly as typed. Use the regex replace tool for patterns. ### Fuzzy match Find the candidates that are most similar to a query, ranked with a score from 0 to 1. - Endpoint: `POST https://setget.net/v1/tools/fuzzy-match` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `fuzzy-match` - Page: https://setget.net/tools/fuzzy-match #### Inputs - `query` (required, string): The text to look for. - `candidates` (required, array): The strings to search, as a list. - `limit` (optional, integer, default `5`, range 1..100): How many matches to return at most. - `min_score` (optional, number, default `0.3`, range 0..1): Leave out candidates that score below this (0 to 1). - `case_sensitive` (optional, boolean, default `false`): Treat upper and lower case as different. #### Example 1 ```json {"input":{"query":"aple","candidates":["apple","maple","grape","banana"]}} ``` Result: ```json {"matches":[{"candidate":"apple","score":0.8889,"index":0},{"candidate":"maple","score":0.8889,"index":1},{"candidate":"grape","score":0.6667,"index":2}],"best":"apple","count":3} ``` #### Example 2: Only the best match ```json {"input":{"query":"instal","candidates":["uninstall","install","init"],"limit":1}} ``` Result: ```json {"matches":[{"candidate":"install","score":0.9231,"index":1}],"best":"install","count":1} ``` **How is the score calculated?** It is twice the number of matching characters divided by the total length of both strings. 1 is an exact match. **Which candidate wins a tie?** The one that comes first in your list. "index" is its position in the list, counting from 0. ### Levenshtein distance Count the single-character edits needed to turn one string into another, with a similarity ratio. - Endpoint: `POST https://setget.net/v1/tools/levenshtein` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `levenshtein` - Page: https://setget.net/tools/levenshtein #### Inputs - `a` (required, string): The first string. - `b` (required, string): The second string. - `case_sensitive` (optional, boolean, default `true`): Treat upper and lower case as different. #### Example 1 ```json {"input":{"a":"kitten","b":"sitting"}} ``` Result: ```json {"distance":3,"similarity":0.5714,"length_a":6,"length_b":7} ``` #### Example 2: Ignoring case ```json {"input":{"a":"Color","b":"colour","case_sensitive":false}} ``` Result: ```json {"distance":1,"similarity":0.8333,"length_a":5,"length_b":6} ``` **How is similarity calculated?** It is 1 minus the distance divided by the length of the longer string, so 1 means identical and 0 means nothing in common. **What counts as one edit?** Inserting, deleting or replacing a single character. Swapping two neighbouring characters counts as two edits. ### Line numbers Add line numbers to text with your choice of start, padding and separator, or strip them off again. - Endpoint: `POST https://setget.net/v1/tools/line-numbers` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `line-numbers` - Page: https://setget.net/tools/line-numbers #### Inputs - `text` (required, string): The text to number, or numbered text to clean. - `action` (optional, string, one of `add`, `remove`, default `"add"`): Add numbers or remove existing ones. - `start` (optional, integer, default `1`, range 0..1000000000): Number given to the first line. - `padding` (optional, string, one of `none`, `space`, `zero`, default `"none"`): Make all numbers the same width with leading spaces or zeros. - `separator` (optional, string, default `": "`): Text placed between the number and the line. Used when adding. - `skip_blank` (optional, boolean, default `false`): Leave blank lines without a number. #### Example 1 ```json {"input":{"text":"alpha\nbeta\ngamma"}} ``` Result: ```json {"text":"1: alpha\n2: beta\n3: gamma","lines":3} ``` #### Example 2: Zero-padded ```json {"input":{"text":"a\nb","start":9,"padding":"zero","separator":" | "}} ``` Result: ```json {"text":"09 | a\n10 | b","lines":2} ``` #### Example 3: Removing numbers ```json {"input":{"text":"1. alpha\n2. beta","action":"remove"}} ``` Result: ```json {"text":"alpha\nbeta","lines":2} ``` **Which number formats does "remove" understand?** A number at the start of the line, with optional spaces before it, followed by a space, a tab or one of . : ) ] | > - **Will "remove" delete numbers that belong to the text?** Yes, if a line starts with one. "2024 was a good year" loses its "2024". Check the result when lines can start with real numbers. ### Lorem ipsum generator Generate lorem ipsum placeholder text by paragraphs, sentences or words. The same request always gives the same text. - Endpoint: `POST https://setget.net/v1/tools/lorem-ipsum` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `lorem-ipsum` - Page: https://setget.net/tools/lorem-ipsum #### Inputs - `units` (optional, string, one of `paragraphs`, `sentences`, `words`, default `"paragraphs"`): What to count in. - `count` (optional, integer, default `3`, range 1..2000): How many to generate: up to 100 paragraphs, 500 sentences or 2,000 words. #### Example 1 ```json {"input":{"units":"words","count":8}} ``` Result: ```json {"text":"Lorem ipsum dolor sit amet consectetur adipiscing elit.","words":8} ``` #### Example 2: Two sentences ```json {"input":{"units":"sentences","count":2}} ``` Result: ```json {"text":"Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.","words":19} ``` **Is the text random?** No. It is built from a fixed passage, so the same units and count always return exactly the same text. That keeps tests and screenshots stable. **How are paragraphs separated?** By one blank line. Each paragraph has five sentences. ### Markdown to plain text Remove Markdown formatting and keep the readable text: headings, emphasis, links, images, lists and code. - Endpoint: `POST https://setget.net/v1/tools/strip-markdown` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `strip-markdown` - Page: https://setget.net/tools/strip-markdown #### Inputs - `text` (required, string): The Markdown to convert. #### Example 1 ```json {"input":{"text":"# Title\n\nSome **bold** and _italic_ text with a [link](https://example.com) and `code`.\n\n- item one\n- item two\n\n![logo](logo.png)"}} ``` Result: ```json {"text":"Title\n\nSome bold and italic text with a link and code.\n\nitem one\nitem two\n\nlogo"} ``` #### Example 2: Quotes, code fences and numbered lists ```json {"input":{"text":"> Quoted **text**\n\n```python\nprint(\"**not bold**\")\n```\n\n1. First\n2. Second"}} ``` Result: ```json {"text":"Quoted text\n\nprint(\"**not bold**\")\n\n1. First\n2. Second"} ``` **What happens to links and images?** A link keeps its text and loses its address. An image is replaced by its alt text. **Is code changed?** No. Text inside code fences and backticks is kept exactly; only the fence lines and backticks are removed. **What happens to lists and tables?** Bullet markers are removed and numbered items keep their numbers. Table rows stay as they are, without the row of dashes under the header. ### PII redactor Replace emails, phone numbers, card numbers, IP addresses and IBANs in text with placeholders such as [EMAIL]. - Endpoint: `POST https://setget.net/v1/tools/redact-pii` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `redact-pii` - Page: https://setget.net/tools/redact-pii #### Inputs - `text` (required, string): The text to redact. - `types` (optional, array): Which kinds to redact: email, phone, card, ip, iban. All of them when omitted. #### Example 1 ```json {"input":{"text":"Email sam@example.com or call +44 20 7946 0958. Card 4111 1111 1111 1111, server 192.168.1.10."}} ``` Result: ```json {"text":"Email [EMAIL] or call [PHONE]. Card [CARD], server [IP].","counts":{"email":1,"phone":1,"card":1,"ip":1,"iban":0},"total":4} ``` #### Example 2: Only bank account numbers ```json {"input":{"text":"Pay GB82 WEST 1234 5698 7654 32 and write to sam@example.com","types":["iban"]}} ``` Result: ```json {"text":"Pay [IBAN] and write to sam@example.com","counts":{"email":0,"phone":0,"card":0,"ip":0,"iban":1},"total":1} ``` **Will it catch every piece of personal data?** No. It is pattern-based and can miss things. It does not recognise names, postal addresses, dates of birth, national ID numbers or unusually written numbers. Always review the result before sharing it. **Can it redact too much?** Yes. Any run of 8 to 15 digits can look like a phone number, so order numbers and other long IDs may be replaced with [PHONE]. **How are card numbers and IBANs recognised?** Card numbers are 13 to 19 digits that pass the Luhn check. IBANs must be in capitals and pass the standard checksum, so mistyped ones are left alone. **Is my text stored?** It is processed in memory to produce the result and not kept. Even so, avoid sending data you are not allowed to share. ### Readability score Score English text with Flesch reading ease and Flesch-Kincaid grade, with a plain-language verdict. - Endpoint: `POST https://setget.net/v1/tools/readability-score` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `readability-score` - Page: https://setget.net/tools/readability-score #### Inputs - `text` (required, string): The English text to score. A few sentences or more gives a steadier result. #### Example ```json {"input":{"text":"The cat sat on the mat. It was a sunny day and the birds were singing."}} ``` Result: ```json {"flesch_reading_ease":103.5,"flesch_kincaid_grade":0.8,"sentences":2,"words":16,"syllables":18,"words_per_sentence":8.0,"syllables_per_word":1.12,"verdict":"Very easy to read. Easily understood by an average 11-year-old."} ``` **What do the numbers mean?** Reading ease usually falls between 0 and 100, and higher is easier; 60 to 70 is plain English. The grade is the US school year needed to follow the text; add 5 for a rough reader age. **Does it work for other languages?** No. The formulas and the syllable counting are made for English, so scores for other languages are not meaningful. **How exact is the syllable count?** It is estimated from spelling and is right for most words, not all. Scores for very short texts can swing a lot. ### Regex extractor Pull every match of a regular expression out of text as a simple list, or just one capture group. - Endpoint: `POST https://setget.net/v1/tools/regex-extract` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `regex-extract` - Page: https://setget.net/tools/regex-extract #### Inputs - `pattern` (required, string): The regular expression (Python syntax). - `text` (required, string): The text to search. - `flags` (optional, string, default `""`): Letters that switch options on: i (ignore case), m (^ and $ match at each line), s (dot matches line breaks), x (allow spaces and comments). - `group` (optional, string, default `"0"`): Capture group to return, by number or name. 0 is the whole match. - `unique` (optional, boolean, default `false`): Return each distinct value once, in order of first appearance. #### Example 1 ```json {"input":{"pattern":"#(\\d+)","text":"Fix #123 and #456, see #123","group":"1","unique":true}} ``` Result: ```json {"matches":["123","456"],"count":2,"truncated":false} ``` #### Example 2: Whole matches ```json {"input":{"pattern":"\\b\\w+@\\w+\\.\\w+","text":"sam@example.com, alex@example.org"}} ``` Result: ```json {"matches":["sam@example.com","alex@example.org"],"count":2,"truncated":false} ``` **Which regex syntax is this?** Python's. It is close to PCRE and JavaScript, with named groups written (?P...). **Is there a limit?** The first 10,000 matches are returned. "truncated" is true when there were more. ### Regex replace Replace every match of a regular expression, with backreferences such as \1 in the replacement. - Endpoint: `POST https://setget.net/v1/tools/regex-replace` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `regex-replace` - Page: https://setget.net/tools/regex-replace #### Inputs - `pattern` (required, string): The regular expression (Python syntax). - `text` (required, string): The text to change. - `replacement` (optional, string, default `""`): Replacement text. \1 or \g inserts a group, \g<0> the whole match. - `flags` (optional, string, default `""`): Letters that switch options on: i (ignore case), m (^ and $ match at each line), s (dot matches line breaks), x (allow spaces and comments). - `count` (optional, integer, default `0`, range 0..): Replace at most this many matches. 0 replaces all. #### Example 1 ```json {"input":{"pattern":"(\\w+)@(\\w+)\\.com","text":"sam@example.com","replacement":"\\2: \\1"}} ``` Result: ```json {"text":"example: sam","replacements":1} ``` #### Example 2: Collapse whitespace ```json {"input":{"pattern":"\\s+","text":"a b \n c","replacement":" "}} ``` Result: ```json {"text":"a b c","replacements":2} ``` **Which regex syntax is this?** Python's. It is close to PCRE and JavaScript, with named groups written (?P...). **Can I write $1 for a group?** No. Use \1, or \g<1> when a digit follows, and \g for named groups. A dollar sign is inserted as it is. **How do I insert a real backslash?** Write two backslashes in the replacement. ### Regex tester Test a regular expression against text and see every match with its position and groups. - Endpoint: `POST https://setget.net/v1/tools/regex-test` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `regex-test` - Page: https://setget.net/tools/regex-test #### Inputs - `pattern` (required, string): The regular expression (Python syntax). - `text` (required, string): The text to search. - `flags` (optional, string, default `""`): Letters that switch options on: i (ignore case), m (^ and $ match at each line), s (dot matches line breaks), x (allow spaces and comments). #### Example 1 ```json {"input":{"pattern":"(?P\\d{4})-(\\d{2})","text":"From 2024-01 to 2025-12"}} ``` Result: ```json {"matched":true,"count":2,"truncated":false,"matches":[{"match":"2024-01","start":5,"end":12,"groups":["2024","01"],"named":{"year":"2024"}},{"match":"2025-12","start":16,"end":23,"groups":["2025","12"],"named":{"year":"2025"}}]} ``` #### Example 2: Ignoring case ```json {"input":{"pattern":"^hello","text":"Hello there","flags":"i"}} ``` Result: ```json {"matched":true,"count":1,"truncated":false,"matches":[{"match":"Hello","start":0,"end":5,"groups":[],"named":{}}]} ``` **Which regex syntax is this?** Python's. It is close to PCRE and JavaScript, with named groups written (?P...). **How are positions counted?** "start" and "end" count characters from 0. "end" is the position just after the match. **Why is a group null?** That group is optional and took no part in the match. ### Remove duplicate lines Remove repeated lines from text, keeping the first occurrence of each and the original order. - Endpoint: `POST https://setget.net/v1/tools/dedupe-lines` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `dedupe-lines` - Page: https://setget.net/tools/dedupe-lines #### Inputs - `text` (required, string): The lines to clean up. - `case_sensitive` (optional, boolean, default `true`): Treat upper and lower case as different lines. - `trim` (optional, boolean, default `false`): Ignore spaces at the start and end of lines when comparing. - `keep_blank` (optional, boolean, default `false`): Keep every blank line instead of treating blank lines as duplicates. #### Example 1 ```json {"input":{"text":"a\nb\na\nc\nb"}} ``` Result: ```json {"text":"a\nb\nc","lines":3,"removed":2} ``` #### Example 2: Ignoring case and stray spaces ```json {"input":{"text":"Sam\nsam \nAlex","case_sensitive":false,"trim":true}} ``` Result: ```json {"text":"Sam\nAlex","lines":2,"removed":1} ``` **Do the lines need to be sorted first?** No. Duplicates are found anywhere in the text, and the remaining lines keep their order. **Which copy is kept?** The first one, exactly as it was written. ### Reverse text Reverse text by characters, by words or by lines. Accents and emoji stay attached to their letters. - Endpoint: `POST https://setget.net/v1/tools/reverse-text` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `reverse-text` - Page: https://setget.net/tools/reverse-text #### Inputs - `text` (required, string): The text to reverse. - `by` (optional, string, one of `characters`, `words`, `lines`, default `"characters"`): What to reverse: the characters, the order of words, or the order of lines. #### Example 1 ```json {"input":{"text":"Hello, world"}} ``` Result: ```json {"text":"dlrow ,olleH"} ``` #### Example 2: Word order ```json {"input":{"text":"one two three","by":"words"}} ``` Result: ```json {"text":"three two one"} ``` #### Example 3: Line order ```json {"input":{"text":"first\nsecond\nthird","by":"lines"}} ``` Result: ```json {"text":"third\nsecond\nfirst"} ``` **Does it break accented letters or emoji?** No. A letter keeps its combining accents, and emoji made of several parts (skin tones, flags, joined emoji) stay whole. ### Slug generator Turn a title into a clean URL slug: accents removed, punctuation dropped, words joined by hyphens. - Endpoint: `POST https://setget.net/v1/tools/slugify` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `slugify` - Page: https://setget.net/tools/slugify #### Inputs - `text` (required, string): The title or phrase to turn into a slug. - `separator` (optional, string, one of `-`, `_`, `.`, default `"-"`): Character placed between words. - `lowercase` (optional, boolean, default `true`): Convert the slug to lower case. - `max_length` (optional, integer, default `0`, range 0..10000): Longest slug allowed, cut at a word boundary. 0 means no limit. #### Example 1 ```json {"input":{"text":"Crème Brûlée: The Ultimate Guide!"}} ``` Result: ```json {"text":"creme-brulee-the-ultimate-guide"} ``` #### Example 2: Cut at a word boundary ```json {"input":{"text":"Hello, World & Friends","separator":"_","max_length":12}} ``` Result: ```json {"text":"hello_world"} ``` **What happens to accented letters?** They are replaced by their plain letter: "é" becomes "e", "ø" becomes "o" and "ß" becomes "ss". **What about non-Latin scripts such as Cyrillic or Chinese?** They are dropped, because they have no agreed plain-letter form. A title written only in such a script gives an empty slug. ### Sort lines Sort lines alphabetically, naturally (file2 before file10), by number or by length, or reverse their order. - Endpoint: `POST https://setget.net/v1/tools/sort-lines` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `sort-lines` - Page: https://setget.net/tools/sort-lines #### Inputs - `text` (required, string): The lines to sort. - `order` (optional, string, one of `alphabetical`, `natural`, `numeric`, `length`, `reverse`, default `"alphabetical"`): How to order the lines. "natural" sorts numbers inside text by value; "reverse" just flips the current order. - `descending` (optional, boolean, default `false`): Sort from last to first (Z to A, largest to smallest). - `case_sensitive` (optional, boolean, default `false`): Treat upper and lower case as different. Capitals then sort before lower case. - `unique` (optional, boolean, default `false`): Remove duplicate lines before sorting. #### Example 1 ```json {"input":{"text":"file10\nfile2\nfile1","order":"natural"}} ``` Result: ```json {"text":"file1\nfile2\nfile10","lines":3,"removed":0} ``` #### Example 2: Alphabetical, ignoring case, without duplicates ```json {"input":{"text":"banana\nApple\ncherry\napple","unique":true}} ``` Result: ```json {"text":"Apple\nbanana\ncherry","lines":3,"removed":1} ``` #### Example 3: By leading number, largest first ```json {"input":{"text":"10 apples\n9 pears\n100 plums","order":"numeric","descending":true}} ``` Result: ```json {"text":"100 plums\n10 apples\n9 pears","lines":3,"removed":0} ``` **What is natural order?** Numbers inside the text are compared by value, so "file2" comes before "file10". Plain alphabetical order puts "file10" first. **How does numeric order treat lines without a number?** Lines are ordered by the number they start with. Lines that do not start with a number go last, in their original order. **Is the result always the same?** Yes. Lines that compare as equal keep the order they had in the input. ### Template renderer Fill a Mustache-style template with JSON data: {{name}} values, {{#list}} loops and {{^empty}} fallbacks. - Endpoint: `POST https://setget.net/v1/tools/template-render` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `template-render` - Page: https://setget.net/tools/template-render #### Inputs - `template` (required, string): The template text. - `data` (required, json): The values to fill in, as a JSON object. - `escape` (optional, boolean, default `true`): HTML-escape {{name}} values. {{{name}}} is never escaped. #### Example 1 ```json {"input":{"template":"Hello {{name}}!\n{{#items}}\n- {{title}} ({{price}})\n{{/items}}\n{{^items}}\nNothing here.\n{{/items}}","data":{"name":"Sam & Co","items":[{"title":"Tea","price":3},{"title":"Cake","price":4.5}]}}} ``` Result: ```json {"text":"Hello Sam & Co!\n- Tea (3)\n- Cake (4.5)\n"} ``` #### Example 2: Dotted names, raw output and a comment ```json {"input":{"template":"{{user.name}} wrote {{{html}}}{{! not shown }}","data":{"user":{"name":"Ada"},"html":"3"}}} ``` Result: ```json {"text":"Ada wrote 3"} ``` #### Example 3: A list of plain values ```json {"input":{"template":"{{#tags}}[{{.}}]{{/tags}}","data":{"tags":["a","b"]}}} ``` Result: ```json {"text":"[a][b]"} ``` **Which Mustache features work?** Variables, {{{raw}}} and {{&raw}} output, dotted names, sections over lists and true values, inverted sections and comments. Partials, lambdas and custom delimiters are not supported. **What counts as false for a section?** false, null, 0, empty text, an empty list, an empty object and a missing name. {{^name}} shows its content in exactly those cases. **What happens when a name is missing from the data?** Nothing is printed for it. No error is raised. **Why did "&" turn into "&"?** Values are HTML-escaped by default. Set escape to false for plain text, or use {{{name}}} for one value. ### Text chunker Split long text into chunks of a maximum size for search and RAG pipelines, breaking at natural boundaries. - Endpoint: `POST https://setget.net/v1/tools/text-chunk` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `text-chunk` - Page: https://setget.net/tools/text-chunk #### Inputs - `text` (required, string): The text to split. - `max_chars` (optional, integer, default `1000`, range 1..1000000): Longest chunk allowed, in characters. - `overlap` (optional, integer, default `0`, range 0..500000): Characters repeated from the end of one chunk at the start of the next. At most half of max_chars. #### Example 1 ```json {"input":{"text":"First paragraph here.\n\nSecond paragraph is a little longer. It has two sentences.","max_chars":40}} ``` Result: ```json {"count":3,"chunks":[{"index":0,"start":0,"end":21,"text":"First paragraph here."},{"index":1,"start":23,"end":59,"text":"Second paragraph is a little longer."},{"index":2,"start":60,"end":81,"text":"It has two sentences."}]} ``` #### Example 2: With overlap ```json {"input":{"text":"alpha beta gamma delta epsilon zeta","max_chars":16,"overlap":6}} ``` Result: ```json {"count":4,"chunks":[{"index":0,"start":0,"end":16,"text":"alpha beta gamma"},{"index":1,"start":11,"end":22,"text":"gamma delta"},{"index":2,"start":17,"end":30,"text":"delta epsilon"},{"index":3,"start":31,"end":35,"text":"zeta"}]} ``` **Where does it break?** At the last paragraph break in the second half of the chunk. If there is none, at the last sentence end, then at the last space, and only then in the middle of a word. **What are start and end?** Character positions in the original text, counting from 0. "end" is the position just after the chunk, and the chunk has no spaces or line breaks at its edges. **How do I size chunks in tokens?** As a rule of thumb one token is about 4 characters of English, so 2,000 characters is roughly 500 tokens. **Is there a limit on the number of chunks?** Yes, 10,000. Raise max_chars if you reach it. ### Text diff Compare two texts line by line and get a unified diff with counts of added and removed lines. - Endpoint: `POST https://setget.net/v1/tools/diff` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `diff` - Page: https://setget.net/tools/diff #### Inputs - `before` (required, string): The original text. - `after` (required, string): The changed text. - `context` (optional, integer, default `3`, range 0..1000): Unchanged lines to show around each change. #### Example 1 ```json {"input":{"before":"apple\nbanana\ncherry\n","after":"apple\nblueberry\ncherry\ndate\n"}} ``` Result: ```json {"text":"--- before\n+++ after\n@@ -1,3 +1,4 @@\n apple\n-banana\n+blueberry\n cherry\n+date\n","added":2,"removed":1,"hunks":1,"equal":false} ``` #### Example 2: Identical texts ```json {"input":{"before":"same\n","after":"same\n"}} ``` Result: ```json {"text":"","added":0,"removed":0,"hunks":0,"equal":true} ``` **What does "\ No newline at end of file" mean?** The line above it is the last line of a text that does not end with a line break. The marker lets a patch tool rebuild the text exactly. **Can I apply the result with patch or git apply?** Yes. The output is a standard unified diff. The file names are "before" and "after". **Are Windows line endings a difference?** No. Windows (CRLF) line endings are treated as plain line breaks in both texts. ### Token estimator Estimate how many LLM tokens a text uses, from its characters and words. An approximation, not an exact count. - Endpoint: `POST https://setget.net/v1/tools/token-estimate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `token-estimate` - Page: https://setget.net/tools/token-estimate #### Inputs - `text` (required, string): The text to estimate. #### Example ```json {"input":{"text":"The quick brown fox jumps over the lazy dog."}} ``` Result: ```json {"tokens":12,"tokens_by_characters":11,"tokens_by_words":12,"characters":44,"words":9,"note":"Approximation: about 4 characters or 0.75 words per token. \"tokens\" is the mean of the two estimates."} ``` **How accurate is this?** It is a rule of thumb: about 4 characters or 0.75 words per token for English prose. Real counts depend on the model and are often 10 to 30 per cent different. **When is the estimate furthest off?** For code, numbers, emoji and languages other than English, which usually need more tokens than this suggests. Use the model's own tokenizer when the exact number matters. ### Truncate text Shorten long text to a maximum length, cutting from the middle, start or end and marking the cut with an ellipsis. - Endpoint: `POST https://setget.net/v1/tools/truncate-middle` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `truncate-middle` - Page: https://setget.net/tools/truncate-middle #### Inputs - `text` (required, string): The text to shorten. - `max_length` (optional, integer, default `80`, range 1..1000000): Longest result allowed, in characters, including the marker. - `where` (optional, string, one of `start`, `middle`, `end`, default `"middle"`): Which part to cut out. "middle" keeps the start and the end. - `marker` (optional, string, default `"\u2026"`): Text that stands in for the removed part. #### Example 1 ```json {"input":{"text":"The quick brown fox jumps over the lazy dog","max_length":20}} ``` Result: ```json {"text":"The quick … lazy dog","truncated":true,"original_length":43,"removed":24} ``` #### Example 2: Keeping the end of a path ```json {"input":{"text":"/home/sam/projects/website/assets/images/logo.png","max_length":26,"where":"start","marker":"..."}} ``` Result: ```json {"text":".../assets/images/logo.png","truncated":true,"original_length":49,"removed":26} ``` #### Example 3: Already short enough ```json {"input":{"text":"Short text","max_length":20}} ``` Result: ```json {"text":"Short text","truncated":false,"original_length":10,"removed":0} ``` **Is the marker counted in the length?** Yes. The result, marker included, is never longer than max_length. **Does it cut accented letters in half?** No. A cut never separates a letter from its combining accents, so the result can be a character or two shorter than the limit. ### Unicode inspector List every character in a text with its code point, name, category and UTF-8 bytes, flagging invisible ones. - Endpoint: `POST https://setget.net/v1/tools/unicode-inspect` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `unicode-inspect` - Page: https://setget.net/tools/unicode-inspect #### Inputs - `text` (required, string): The text to inspect. The first 200 characters are listed. #### Example 1 ```json {"input":{"text":"a​é"}} ``` Result: ```json {"count":3,"flagged":1,"truncated":false,"characters":[{"char":"a","codepoint":"U+0061","name":"LATIN SMALL LETTER A","category":"Ll","utf8":"61","warning":null},{"char":"​","codepoint":"U+200B","name":"ZERO WIDTH SPACE","category":"Cf","utf8":"E2 80 8B","warning":"Invisible zero-width character."},{"char":"é","codepoint":"U+00E9","name":"LATIN SMALL LETTER E WITH ACUTE","category":"Ll","utf8":"C3 A9","warning":null}]} ``` #### Example 2: A Cyrillic letter that looks Latin ```json {"input":{"text":"рa"}} ``` Result: ```json {"count":2,"flagged":1,"truncated":false,"characters":[{"char":"р","codepoint":"U+0440","name":"CYRILLIC SMALL LETTER ER","category":"Ll","utf8":"D1 80","warning":"Looks like the Latin letter \"p\" but is a different character."},{"char":"a","codepoint":"U+0061","name":"LATIN SMALL LETTER A","category":"Ll","utf8":"61","warning":null}]} ``` **What gets flagged?** Zero-width and text-direction characters, non-breaking and other unusual spaces, control characters, curly quotes, dash look-alikes, full-width forms and common Cyrillic or Greek letters that look like Latin ones. **What do the category codes mean?** They are Unicode general categories: Lu and Ll are upper and lower case letters, Nd is a digit, Zs a space, Cc a control character, Cf an invisible formatting character, Mn a combining mark. **Is the look-alike check complete?** No. It covers the most common look-alike letters, not every confusable character in Unicode. **How do I remove the flagged characters?** Use the Unicode normaliser with strip_control switched on. ### Unicode normaliser Normalise text to NFC, NFD, NFKC or NFKD, and optionally strip accents or invisible characters. - Endpoint: `POST https://setget.net/v1/tools/unicode-normalize` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `unicode-normalize` - Page: https://setget.net/tools/unicode-normalize #### Inputs - `text` (required, string): The text to normalise. - `form` (optional, string, one of `NFC`, `NFD`, `NFKC`, `NFKD`, default `"NFC"`): Normal form. NFC is the usual choice; the K forms also replace look-alike variants such as "fi" and full-width letters. - `strip_accents` (optional, boolean, default `false`): Remove accents and other marks from letters. - `strip_control` (optional, boolean, default `false`): Remove zero-width, text-direction and control characters. Tabs and line breaks stay. #### Example 1 ```json {"input":{"text":"café"}} ``` Result: ```json {"text":"café","changed":true,"characters_before":5,"characters_after":4} ``` #### Example 2: Removing accents ```json {"input":{"text":"Crème brûlée","strip_accents":true}} ``` Result: ```json {"text":"Creme brulee","changed":true,"characters_before":12,"characters_after":12} ``` #### Example 3: Removing zero-width characters ```json {"input":{"text":"a​b","strip_control":true}} ``` Result: ```json {"text":"ab","changed":true,"characters_before":4,"characters_after":2} ``` **Which form should I use?** NFC for storing and comparing text. NFKC when look-alike variants (full-width letters, ligatures, superscripts) should become ordinary characters. **Why do two strings look the same but compare as different?** One may use a single accented character and the other a letter followed by a combining accent. Normalising both to the same form fixes that. **Can strip_control change emoji?** Yes. Emoji such as family groups are joined by an invisible zero-width joiner, so removing it splits them into separate emoji. ### Word and character counter Count characters, words, lines, sentences, paragraphs and UTF-8 bytes in a text. - Endpoint: `POST https://setget.net/v1/tools/text-count` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `text-count` - Page: https://setget.net/tools/text-count #### Inputs - `text` (required, string): The text to count. #### Example 1 ```json {"input":{"text":"Hello world. How are you?\n\nFine, thanks!"}} ``` Result: ```json {"characters":40,"characters_no_spaces":33,"words":7,"lines":3,"sentences":3,"paragraphs":2,"bytes":40} ``` #### Example 2: Accented letters take two bytes ```json {"input":{"text":"naïve café"}} ``` Result: ```json {"characters":10,"characters_no_spaces":9,"words":2,"lines":1,"sentences":1,"paragraphs":1,"bytes":12} ``` **Why are characters and bytes different?** Letters outside basic English take two to four bytes in UTF-8. An emoji is usually one or two characters but four or more bytes. **What is a word here?** Any run of characters between spaces or line breaks. "well-known" is one word. **How are sentences counted?** By full stops, question marks and exclamation marks followed by a space. Abbreviations such as "Dr. Smith" can add an extra sentence. ### Word frequency counter Count how often each word appears in a text and list the most common ones, optionally skipping stop words. - Endpoint: `POST https://setget.net/v1/tools/word-frequency` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `word-frequency` - Page: https://setget.net/tools/word-frequency #### Inputs - `text` (required, string): The text to analyse. - `top` (optional, integer, default `10`, range 1..1000): How many words to list. - `min_length` (optional, integer, default `1`, range 1..100): Ignore words shorter than this many characters. - `case_sensitive` (optional, boolean, default `false`): Count "Word" and "word" separately. - `ignore_stop_words` (optional, boolean, default `false`): Leave out common English words such as "the", "and" and "of". #### Example 1 ```json {"input":{"text":"the cat and the hat and the bat","top":2}} ``` Result: ```json {"words":[{"word":"the","count":3},{"word":"and","count":2}],"total_words":8,"unique_words":5} ``` #### Example 2: Without stop words ```json {"input":{"text":"The cat and the hat and the bat","ignore_stop_words":true}} ``` Result: ```json {"words":[{"word":"cat","count":1},{"word":"hat","count":1},{"word":"bat","count":1}],"total_words":3,"unique_words":3} ``` **What counts as a word?** A run of letters or digits. Apostrophes inside a word are kept, so "don't" is one word; "well-known" counts as two. **How are ties ordered?** Words with the same count are listed in the order they first appear in the text. **What do the totals cover?** total_words and unique_words count the words left after the min_length and stop-word filters. ### Wrap and indent text Hard-wrap text to a line width, add or remove indentation, and put a prefix such as "> " on every line. - Endpoint: `POST https://setget.net/v1/tools/wrap-indent` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `wrap-indent` - Page: https://setget.net/tools/wrap-indent #### Inputs - `text` (required, string): The text to wrap or indent. - `width` (optional, integer, default `80`, range 0..10000): Longest line allowed, counting the indent and prefix. 0 leaves line lengths alone. - `indent` (optional, integer, default `0`, range -100..100): Spaces to add to every line. A negative number removes up to that many leading spaces. - `prefix` (optional, string, default `""`): Text put at the start of every line, for example "> " or "# ". - `reflow` (optional, boolean, default `false`): Join the lines of each paragraph before wrapping, so short lines are filled up. - `dedent` (optional, boolean, default `false`): First remove the indentation that all lines have in common. #### Example 1 ```json {"input":{"text":"The quick brown fox jumps over the lazy dog.","width":20,"prefix":"> "}} ``` Result: ```json {"text":"> The quick brown\n> fox jumps over the\n> lazy dog.","lines":3} ``` #### Example 2: Indent only ```json {"input":{"text":"def f():\n return 1","width":0,"indent":4}} ``` Result: ```json {"text":" def f():\n return 1","lines":2} ``` #### Example 3: Reflow paragraphs ```json {"input":{"text":"one\ntwo\nthree\n\nfour","width":12,"reflow":true}} ``` Result: ```json {"text":"one two\nthree\n\nfour","lines":4} ``` **Are long words split?** Only when a single word is longer than the width. Otherwise lines break at spaces. **What happens to a line's own indentation?** It is kept, and the continuation lines created by wrapping get the same indentation. **How is width measured?** In characters. Wide characters such as Chinese or emoji count as one, so they may look wider on screen. ## Encoding and IDs Encode and decode, hash, compress, and make IDs, passwords and random values. ### All hashes at once Compute every supported hash of one input at once: MD5, SHA-1, SHA-2, SHA3-256 and BLAKE2b. - Endpoint: `POST https://setget.net/v1/tools/hash-all` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `hash-all` - Page: https://setget.net/tools/hash-all #### Inputs - `text` (required, string): The input to hash. - `input_encoding` (optional, string, one of `text`, `hex`, `base64`, default `"text"`): How the input is written: plain text (read as UTF-8), hex digits or base64. - `output` (optional, string, one of `hex`, `base64`, default `"hex"`): How to write the result. #### Example ```json {"input":{"text":"abc"}} ``` Result: ```json {"hashes":{"md5":"900150983cd24fb0d6963f7d28e17f72","sha1":"a9993e364706816aba3e25717850c26c9cd0d89d","sha256":"ba7816bf8f01cfea414140de5dae2223b00361a396177a9cb410ff61f20015ad","sha384":"cb00753f45a35e8bb5a03d699ac65007272c32ab0eded1631a8b605a43ff5bed8086072ba1e7cc2358baeca134c825a7","sha512":"ddaf35a193617abacc417349ae20413112e6fa4e89a97ea20a9eeee64b55d39a2192992a274fc1a836ba3c23a3feebbd454d4423643ce80e2a9ac94fa54ca49f","sha3-256":"3a985da74fe225b2045c172d6bd390bd855f086e3e9d525b46bfe24511431532","blake2b":"ba80a53f981c4d0d6a2797b69f12f6e94c212f14685ac4b74b12bb6fdbffa2d17d87c5392aab792dc252d5de4533cc9518d38aa8dbf1925ab92386edd4009923"},"bytes":3} ``` **Are MD5 and SHA-1 safe to use?** Not for security. Both are broken for signatures and certificates, and no plain hash, however strong, is suitable for storing passwords: use a password hashing function such as Argon2, scrypt or bcrypt. MD5 and SHA-1 are still fine for spotting accidental file corruption. **How do I find which algorithm made a hash?** Hash the original input here and look for the matching line. The length is also a clue: 32 hex digits is MD5, 40 is SHA-1 and 64 is SHA-256 or SHA3-256. ### Base32 encoder and decoder Encode text to Base32 (RFC 4648) or decode it back, with or without padding. - Endpoint: `POST https://setget.net/v1/tools/base32` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `base32` - Page: https://setget.net/tools/base32 #### Inputs - `text` (required, string): The text to encode, or the Base32 to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text to Base32, or decode Base32 to text. - `padding` (optional, boolean, default `true`): Keep the trailing "=" padding when encoding. #### Example 1 ```json {"input":{"text":"Hello"}} ``` Result: ```json {"text":"JBSWY3DP","bytes":5} ``` #### Example 2: Padding fills the last group ```json {"input":{"text":"foobar"}} ``` Result: ```json {"text":"MZXW6YTBOI======","bytes":6} ``` #### Example 3: Lower case, spaces and no padding ```json {"input":{"text":"mzxw 6ytb oi","action":"decode"}} ``` Result: ```json {"text":"foobar","encoding":"utf-8","binary":false,"bytes":6} ``` **Which alphabet is used?** The standard RFC 4648 alphabet: A-Z and 2-7. Decoding ignores case, spaces and missing padding. **What if the decoded data is not text?** Bytes that are not valid UTF-8 are returned as hex, with "binary" set to true. ### Base58 encoder and decoder Encode text or bytes to Base58 with the Bitcoin alphabet, or decode Base58 back. - Endpoint: `POST https://setget.net/v1/tools/base58` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `base58` - Page: https://setget.net/tools/base58 #### Inputs - `text` (required, string): The value to encode, or the Base58 to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode to Base58, or decode Base58. - `input_encoding` (optional, string, one of `text`, `hex`, `base64`, default `"text"`): When encoding, how the input is written: plain text (UTF-8), hex digits or base64. #### Example 1 ```json {"input":{"text":"Hello World!"}} ``` Result: ```json {"text":"2NEpo7TZRRrLZSi2U","bytes":12} ``` #### Example 2: Leading zero bytes become leading 1s ```json {"input":{"text":"0000287fb4cd","input_encoding":"hex"}} ``` Result: ```json {"text":"11233QC4","bytes":6} ``` #### Example 3 ```json {"input":{"text":"2NEpo7TZRRrLZSi2U","action":"decode"}} ``` Result: ```json {"text":"Hello World!","encoding":"utf-8","binary":false,"bytes":12} ``` **Is this Base58Check?** No. This is plain Base58. Base58Check, used for Bitcoin addresses, also appends a four-byte checksum, which is not added or verified here. **Why are 0, O, I and l missing?** The Bitcoin alphabet leaves them out because they look alike when read or typed. **What if the decoded data is not text?** Bytes that are not valid UTF-8, such as keys and addresses, are returned as hex with "binary" set to true. ### Base64 encoder and decoder Encode text to Base64 or decode it back, in the standard or URL-safe alphabet. - Endpoint: `POST https://setget.net/v1/tools/base64` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `base64` - Page: https://setget.net/tools/base64 #### Inputs - `text` (required, string): The text to encode, or the Base64 to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text to Base64, or decode Base64 to text. - `variant` (optional, string, one of `standard`, `url-safe`, default `"standard"`): Alphabet for encoding. "url-safe" uses - and _ in place of + and /. Decoding accepts both. - `padding` (optional, boolean, default `true`): Keep the trailing "=" padding when encoding. #### Example 1 ```json {"input":{"text":"Hello, World!"}} ``` Result: ```json {"text":"SGVsbG8sIFdvcmxkIQ==","bytes":13} ``` #### Example 2: Decoding without padding ```json {"input":{"text":"SGVsbG8sIFdvcmxkIQ","action":"decode"}} ``` Result: ```json {"text":"Hello, World!","encoding":"utf-8","binary":false,"bytes":13} ``` #### Example 3: URL-safe, no padding ```json {"input":{"text":"subjects?_d=1","variant":"url-safe","padding":false}} ``` Result: ```json {"text":"c3ViamVjdHM_X2Q9MQ","bytes":13} ``` **Is Base64 encryption?** No. Anyone can decode it. Base64 only makes data safe to carry as text; it hides nothing. **What if the decoded data is not text?** Bytes that are not valid UTF-8 are returned as hex, with "binary" set to true. **Do I need the "=" signs to decode?** No. Missing padding, line breaks and spaces are all accepted. ### CRC-32 and Adler-32 checksum Compute the CRC-32 and Adler-32 checksums of text or bytes, as decimal and hex. - Endpoint: `POST https://setget.net/v1/tools/crc32` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `crc32` - Page: https://setget.net/tools/crc32 #### Inputs - `text` (required, string): The input to checksum. - `input_encoding` (optional, string, one of `text`, `hex`, `base64`, default `"text"`): How the input is written: plain text (read as UTF-8), hex digits or base64. #### Example ```json {"input":{"text":"123456789"}} ``` Result: ```json {"crc32":{"decimal":3421780262,"hex":"cbf43926"},"adler32":{"decimal":152961502,"hex":"091e01de"},"bytes":9} ``` **Which CRC-32 is this?** The common one used by ZIP, gzip, PNG and Ethernet (polynomial 0x04C11DB7, reflected). It is not CRC-32C. **Can I use a CRC for security?** No. Checksums catch accidental damage only. Anyone can craft different data with the same CRC. Use SHA-256 to detect tampering. ### Deflate compress and decompress Compress text to a zlib or raw deflate stream as base64, or decompress one. The format is detected. - Endpoint: `POST https://setget.net/v1/tools/deflate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `deflate` - Page: https://setget.net/tools/deflate #### Inputs - `text` (required, string): The text to compress, or base64 zlib or deflate data to decompress. - `action` (optional, string, one of `compress`, `decompress`, default `"compress"`): Compress text, or decompress deflate data. - `format` (optional, string, one of `zlib`, `raw`, default `"zlib"`): When compressing: "zlib" adds the two-byte header and checksum, "raw" is the bare deflate stream. Decompression detects the format itself. - `level` (optional, integer, default `6`, range 1..9): Compression effort when compressing, from 1 (fastest) to 9 (smallest). #### Example 1 ```json {"input":{"text":"na na na na na na na na na na na na na na na na Batman!"}} ``` Result: ```json {"text":"eJzLS1TIIwU5JZbkJuYpAgDl2hFl","bytes_before":55,"bytes_after":21,"ratio":0.382} ``` #### Example 2: A raw deflate stream ```json {"input":{"text":"na na na na na na na na na na na na na na na na Batman!","format":"raw"}} ``` Result: ```json {"text":"y0tUyCMFOSWW5CbmKQIA","bytes_before":55,"bytes_after":15,"ratio":0.273} ``` #### Example 3: The format is detected ```json {"input":{"text":"eJzLS1TIIwU5JZbkJuYpAgDl2hFl","action":"decompress"}} ``` Result: ```json {"text":"na na na na na na na na na na na na na na na na Batman!","encoding":"utf-8","binary":false,"bytes":55,"format":"zlib"} ``` **What is the difference between zlib, raw deflate and gzip?** All three hold the same deflate-compressed data. Zlib wraps it in a 2-byte header and an Adler-32 checksum, gzip in a larger header and a CRC-32, and raw deflate has no wrapper at all. **Is there a size limit?** Yes. Decompression stops with an error once the output passes 5 MB, which also protects against "zip bomb" data that expands enormously. **What if the decompressed data is not text?** Bytes that are not valid UTF-8 are returned as base64, with "binary" set to true. ### Gzip compress and decompress Compress text with gzip to base64 and see the size saved, or decompress base64 gzip data. - Endpoint: `POST https://setget.net/v1/tools/gzip` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `gzip` - Page: https://setget.net/tools/gzip #### Inputs - `text` (required, string): The text to compress, or base64 gzip data to decompress. - `action` (optional, string, one of `compress`, `decompress`, default `"compress"`): Compress text, or decompress gzip data. - `level` (optional, integer, default `6`, range 1..9): Compression effort when compressing, from 1 (fastest) to 9 (smallest). #### Example 1 ```json {"input":{"text":"na na na na na na na na na na na na na na na na Batman!"}} ``` Result: ```json {"text":"H4sIAAAAAAAA/8tLVMgjBTklluQm5ikCANmXu+U3AAAA","bytes_before":55,"bytes_after":33,"ratio":0.6} ``` #### Example 2: Back to text ```json {"input":{"text":"H4sIAAAAAAAA/8tLVMgjBTklluQm5ikCANmXu+U3AAAA","action":"decompress"}} ``` Result: ```json {"text":"na na na na na na na na na na na na na na na na Batman!","encoding":"utf-8","binary":false,"bytes":55} ``` **Why is the compressed result larger than my text?** Gzip adds 18 bytes of header and checksum, and base64 adds a third on top. Compression only pays off on longer or repetitive text. **Is there a size limit?** Yes. Decompression stops with an error once the output passes 5 MB, which also protects against "zip bomb" data that expands enormously. **What if the decompressed data is not text?** Bytes that are not valid UTF-8 are returned as base64, with "binary" set to true. ### Hash generator Compute an MD5, SHA-1, SHA-256, SHA-384, SHA-512, SHA3-256 or BLAKE2b hash of text or bytes. - Endpoint: `POST https://setget.net/v1/tools/hash` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `hash` - Page: https://setget.net/tools/hash #### Inputs - `text` (required, string): The input to hash. - `algorithm` (optional, string, one of `md5`, `sha1`, `sha256`, `sha384`, `sha512`, `sha3-256`, `blake2b`, default `"sha256"`): The hash function to use. - `input_encoding` (optional, string, one of `text`, `hex`, `base64`, default `"text"`): How the input is written: plain text (read as UTF-8), hex digits or base64. - `output` (optional, string, one of `hex`, `base64`, default `"hex"`): How to write the result. #### Example 1 ```json {"input":{"text":"hello"}} ``` Result: ```json {"text":"2cf24dba5fb0a30e26e83b2ac5b9e29e1b161e5c1fa7425e73043362938b9824","algorithm":"sha256","bits":256,"bytes":5} ``` #### Example 2: MD5 ```json {"input":{"text":"abc","algorithm":"md5"}} ``` Result: ```json {"text":"900150983cd24fb0d6963f7d28e17f72","algorithm":"md5","bits":128,"bytes":3} ``` #### Example 3: Hex input, base64 output ```json {"input":{"text":"616263","input_encoding":"hex","algorithm":"sha1","output":"base64"}} ``` Result: ```json {"text":"qZk+NkcGgWq6PiVxeFDCbJzQ2J0=","algorithm":"sha1","bits":160,"bytes":3} ``` **Are MD5 and SHA-1 safe to use?** Not for security. Both are broken for signatures and certificates, and no plain hash, however strong, is suitable for storing passwords: use a password hashing function such as Argon2, scrypt or bcrypt. MD5 and SHA-1 are still fine for spotting accidental file corruption. **Why is my hash different from another tool?** Usually because of an invisible difference in the input: a trailing line break, a different line ending or a different text encoding. Text is hashed here as UTF-8 exactly as sent. **Can a hash be reversed?** No. A hash is one-way. Short or common inputs can still be guessed by trying candidates, which is why plain hashes must not be used for passwords. ### Hex encoder and decoder Convert text to hexadecimal bytes and back, with optional upper case and separators. - Endpoint: `POST https://setget.net/v1/tools/hex` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `hex` - Page: https://setget.net/tools/hex #### Inputs - `text` (required, string): The text to encode, or the hex to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text to hex, or decode hex to text. - `uppercase` (optional, boolean, default `false`): Write the hex digits A-F in upper case when encoding. - `separator` (optional, string, one of `none`, `space`, `colon`, default `"none"`): What to put between bytes when encoding. #### Example 1 ```json {"input":{"text":"Hello"}} ``` Result: ```json {"text":"48656c6c6f","bytes":5} ``` #### Example 2: Upper case with colons ```json {"input":{"text":"Hi!","uppercase":true,"separator":"colon"}} ``` Result: ```json {"text":"48:69:21","bytes":3} ``` #### Example 3: Decoding ignores prefixes and separators ```json {"input":{"text":"0x48 0x65 6c:6c:6f","action":"decode"}} ``` Result: ```json {"text":"Hello","bytes":5} ``` **Which separators can I decode?** Spaces, line breaks, colons, commas, hyphens and "0x" prefixes are all ignored when decoding. **Why does decoding fail with "not UTF-8"?** The bytes are binary data, not text. Hex is already the clearest way to show them. ### HMAC generator Compute an HMAC signature of a message with a secret key, as hex or base64. - Endpoint: `POST https://setget.net/v1/tools/hmac` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `hmac` - Page: https://setget.net/tools/hmac #### Inputs - `message` (required, string): The message to sign. - `key` (required, string): The secret key. Use a test key here, never a production secret. - `algorithm` (optional, string, one of `md5`, `sha1`, `sha256`, `sha384`, `sha512`, `sha3-256`, `blake2b`, default `"sha256"`): The hash function to use. - `key_encoding` (optional, string, one of `text`, `hex`, `base64`, default `"text"`): How the key is written: plain text (UTF-8), hex digits or base64. - `output` (optional, string, one of `hex`, `base64`, default `"hex"`): How to write the result. #### Example 1 ```json {"input":{"message":"what do ya want for nothing?","key":"Jefe"}} ``` Result: ```json {"text":"5bdcc146bf60754e6a042426089575c75a003f089d2739839dec58b964ec3843","algorithm":"sha256","bits":256} ``` #### Example 2: A hex key with SHA-1 ```json {"input":{"message":"The quick brown fox jumps over the lazy dog","key":"6b6579","key_encoding":"hex","algorithm":"sha1"}} ``` Result: ```json {"text":"de7c9b85b8b78aa6bc8a7a36f70a90701c9db4d9","algorithm":"sha1","bits":160} ``` **Is it safe to enter my real secret key?** No. Never send a production secret to any online tool. Nothing is stored here, but a real key should stay on your own systems. Test with a throwaway key. **Why does my webhook signature not match?** Sign the raw request body exactly as received, byte for byte. Re-formatted JSON, a different line ending or a key given as text when it is really hex will all change the result. **Is HMAC with MD5 or SHA-1 broken?** HMAC holds up better than the plain hashes, but new systems should use SHA-256 or stronger. ### HTML entity encoder and decoder Escape text as HTML entities, named or numeric, or decode entities back to characters. - Endpoint: `POST https://setget.net/v1/tools/html-entities` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `html-entities` - Page: https://setget.net/tools/html-entities #### Inputs - `text` (required, string): The text to escape, or text with entities to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode characters as entities, or decode entities to characters. - `style` (optional, string, one of `named`, `decimal`, `hex`, default `"named"`): How to write entities: named where a name exists (é), decimal (é) or hex (é). - `scope` (optional, string, one of `all`, `special`, `non-ascii`, default `"all"`): What to encode: "special" is only & < > " and ', "non-ascii" is only characters outside plain ASCII, "all" is both. #### Example 1 ```json {"input":{"text":"

Café & \"tea\"

"}} ``` Result: ```json {"text":"<p>Café & "tea"</p>","replaced":8} ``` #### Example 2: Only non-ASCII characters, leaving the markup alone ```json {"input":{"text":"Café €5","style":"decimal","scope":"non-ascii"}} ``` Result: ```json {"text":"Café €5","replaced":2} ``` #### Example 3 ```json {"input":{"text":"<p>Café & €5 😀</p>","action":"decode"}} ``` Result: ```json {"text":"

Café & €5 😀

"} ``` **Which characters must be escaped in HTML?** Always & and <. Inside attribute values also the quote character. Escaping > and both quotes everywhere, as "special" does, is the safe habit. **Is escaping enough to stop XSS?** Only when the text is placed in HTML content or a quoted attribute. Text placed inside a script, a style or a URL needs different escaping. **Why is the apostrophe written as '?** It works everywhere, while the named form ' is not understood by some older HTML readers. ### JWT decoder Decode the header and payload of a JSON Web Token and show its times, without verifying it. - Endpoint: `POST https://setget.net/v1/tools/jwt-decode` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `jwt-decode` - Page: https://setget.net/tools/jwt-decode #### Inputs - `token` (required, string): The JWT to decode. Use a test or expired token, never a live one. - `now` (optional, string): The time to compare "exp" and "nbf" against, as an ISO 8601 time. Adds the "expired" flag. Left out, no comparison is made. #### Example 1 ```json {"input":{"token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzdWIiOiIxMjM0NTY3ODkwIiwibmFtZSI6IkpvaG4gRG9lIiwiaWF0IjoxNTE2MjM5MDIyfQ.SflKxwRJSMeKKF2QT4fwpMeJf36POk6yJV_adQssw5c"}} ``` Result: ```json {"header":{"alg":"HS256","typ":"JWT"},"payload":{"sub":"1234567890","name":"John Doe","iat":1516239022},"signature":"SflKxwRJSMeKKF2QT4fwpMeJf36POk6yJV_adQssw5c","algorithm":"HS256","times":{"iat":"2018-01-18T01:30:22Z"},"verified":false} ``` #### Example 2: Checking expiry against a given time ```json {"input":{"token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzdWIiOiI0MiIsIm5iZiI6MTY5OTk5NjQwMCwiZXhwIjoxNzAwMDAwMDAwfQ.c2lnbmF0dXJl","now":"2024-01-01T00:00:00Z"}} ``` Result: ```json {"header":{"alg":"HS256","typ":"JWT"},"payload":{"sub":"42","nbf":1699996400,"exp":1700000000},"signature":"c2lnbmF0dXJl","algorithm":"HS256","times":{"exp":"2023-11-14T22:13:20Z","nbf":"2023-11-14T21:13:20Z"},"verified":false,"now":"2024-01-01T00:00:00Z","expired":true,"not_yet_valid":false} ``` **Does this verify the signature?** No. It only decodes. A decoded token proves nothing about who made it, so never trust its contents without verifying the signature in your own code. **Is it safe to paste a live token?** No. A valid token works like a password until it expires. Never paste live tokens into any online tool; use an expired or test token. **Is the payload encrypted?** No. The header and payload of a signed JWT are only base64url-encoded, so anyone holding the token can read them. **Why is there no "expired" flag?** The tool never reads the clock. Send the current time as "now" to have "exp" and "nbf" compared against it. ### Morse code translator Translate text to international Morse code and back, with "/" between words. - Endpoint: `POST https://setget.net/v1/tools/morse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `morse` - Page: https://setget.net/tools/morse #### Inputs - `text` (required, string): The text to translate, or Morse code written with dots and dashes. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text to Morse, or decode Morse to text. #### Example 1 ```json {"input":{"text":"SOS"}} ``` Result: ```json {"text":"... --- ...","unknown":[]} ``` #### Example 2: Words are separated by a slash ```json {"input":{"text":"Hello World"}} ``` Result: ```json {"text":".... . .-.. .-.. --- / .-- --- .-. .-.. -..","unknown":[]} ``` #### Example 3 ```json {"input":{"text":".... . .-.. .-.. --- / .-- --- .-. .-.. -..","action":"decode"}} ``` Result: ```json {"text":"HELLO WORLD","unknown":[]} ``` #### Example 4: Characters with no Morse code are reported ```json {"input":{"text":"café #1"}} ``` Result: ```json {"text":"-.-. .- ..-. / .----","unknown":["É","#"]} ``` **How should I write Morse for decoding?** Use "." and "-" with a space between letters and "/" between words. A new line also counts as a word break. **What happens to characters without a Morse code?** They are left out of the result and listed under "unknown". International Morse covers A-Z, 0-9 and common punctuation only. **Why is decoded text in capitals?** Morse code has no upper or lower case, so capitals are the convention. ### Nano ID generator Generate short, URL-safe random IDs of any length, optionally from your own alphabet. - Endpoint: `POST https://setget.net/v1/tools/nanoid` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `nanoid` - Page: https://setget.net/tools/nanoid #### Inputs - `size` (optional, integer, default `21`, range 1..256): Characters per ID. - `count` (optional, integer, default `1`, range 1..100): How many to generate. - `alphabet` (optional, string): Characters to draw from, each listed once. Left out, A-Z, a-z, 0-9, "_" and "-" are used. #### Example 1: One standard 21-character ID ```json {"input":{}} ``` #### Example 2: Five short hex IDs ```json {"input":{"size":10,"count":5,"alphabet":"0123456789abcdef"}} ``` **How likely is a collision?** With the default 21 characters there are 126 random bits, about the same as a random UUID. Shorter IDs or smaller alphabets collide sooner; "entropy_bits" in the result shows the strength. **Is the output evenly random?** Yes. Each character is drawn from the operating system's secure random source without modulo bias. ### Passphrase generator Generate memorable passphrases from random common English words, and see their entropy. - Endpoint: `POST https://setget.net/v1/tools/passphrase-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `passphrase-generate` - Page: https://setget.net/tools/passphrase-generate #### Inputs - `words` (optional, integer, default `5`, range 2..20): Words per passphrase. - `count` (optional, integer, default `1`, range 1..100): How many to generate. - `separator` (optional, string, default `"-"`): Text placed between the words. - `capitalise` (optional, boolean, default `false`): Start each word with a capital letter. - `add_number` (optional, boolean, default `false`): Add a random number from 0 to 99 at the end. #### Example 1: Five random words ```json {"input":{}} ``` #### Example 2: Capitalised, with a number, for sites that demand both ```json {"input":{"words":4,"count":3,"separator":" ","capitalise":true,"add_number":true}} ``` **Are the results stored?** No. They are made with the operating system's secure random source, returned once and not kept. For the highest-value secrets, generate them on your own device with a password manager. **How strong is a passphrase?** Each word adds about 9.5 bits of entropy, so five words give about 48 bits and seven about 67. Use six or more words for anything important; "entropy_bits" shows the exact figure. **Do capitals or a separator make it stronger?** Hardly. They are fixed, not random, so they add nothing to the entropy. Only more words, or the random number, add strength. ### Password generator Generate strong random passwords with the character types you choose, and see their entropy. - Endpoint: `POST https://setget.net/v1/tools/password-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `password-generate` - Page: https://setget.net/tools/password-generate #### Inputs - `length` (optional, integer, default `16`, range 4..256): Characters per password. - `count` (optional, integer, default `1`, range 1..100): How many to generate. - `lower` (optional, boolean, default `true`): Include lower-case letters a-z. - `upper` (optional, boolean, default `true`): Include upper-case letters A-Z. - `digits` (optional, boolean, default `true`): Include the digits 0-9. - `symbols` (optional, boolean, default `true`): Include symbols such as ! @ # and %. - `exclude_similar` (optional, boolean, default `false`): Leave out characters that look alike: I, l, 1, |, O, 0 and o. #### Example 1: One 16-character password ```json {"input":{}} ``` #### Example 2: Five long passwords that are easy to read aloud ```json {"input":{"length":24,"count":5,"symbols":false,"exclude_similar":true}} ``` **Are the results stored?** No. They are made with the operating system's secure random source, returned once and not kept. For the highest-value secrets, generate them on your own device with a password manager. **How many entropy bits are enough?** About 80 bits or more resists offline guessing for a long time. The default 16 characters from all four types give roughly 100 bits. **Does every password contain each chosen type?** Yes. At least one character of every selected type is always included, at a random position. ### Punycode converter Convert internationalised domain names to their ASCII "xn--" form and back. - Endpoint: `POST https://setget.net/v1/tools/punycode` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `punycode` - Page: https://setget.net/tools/punycode #### Inputs - `text` (required, string): The domain name to convert. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode a Unicode domain to ASCII, or decode an "xn--" domain to Unicode. #### Example 1 ```json {"input":{"text":"münchen.de"}} ``` Result: ```json {"text":"xn--mnchen-3ya.de","changed":true} ``` #### Example 2 ```json {"input":{"text":"xn--mnchen-3ya.de","action":"decode"}} ``` Result: ```json {"text":"münchen.de","changed":true} ``` #### Example 3: Every label is converted on its own ```json {"input":{"text":"例え.テスト"}} ``` Result: ```json {"text":"xn--r8jz45g.xn--zckzah","changed":true} ``` **What is the "xn--" prefix?** It marks a domain label that holds non-ASCII characters in encoded form. DNS itself only carries ASCII, so browsers convert names both ways. **Why did "ß" become "ss"?** Encoding follows the older IDNA 2003 rules, which fold a few characters: the German sharp s becomes "ss" and capitals become lower case. Registries using IDNA 2008 keep "ß" as its own character. **Why can look-alike domains be dangerous?** Letters from different scripts can look identical, for example Latin "a" and Cyrillic "а". Decoding an "xn--" name here shows what it really contains. ### Quoted-printable encoder and decoder Encode text as quoted-printable, the "=C3=A9" style used in email, or decode it. - Endpoint: `POST https://setget.net/v1/tools/quoted-printable` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `quoted-printable` - Page: https://setget.net/tools/quoted-printable #### Inputs - `text` (required, string): The text to encode, or quoted-printable text to decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text, or decode quoted-printable. - `charset` (optional, string, one of `utf-8`, `iso-8859-1`, `windows-1252`, default `"utf-8"`): The character set of the bytes. Check the "charset" in the email's Content-Type header. #### Example 1 ```json {"input":{"text":"Café = coffee"}} ``` Result: ```json {"text":"Caf=C3=A9 =3D coffee"} ``` #### Example 2: A soft line break ("=" at the end of a line) is removed ```json {"input":{"text":"Caf=C3=A9 =3D cof=\nfee","action":"decode"}} ``` Result: ```json {"text":"Café = coffee","encoding":"utf-8","binary":false,"bytes":14} ``` #### Example 3: An older Latin-1 email ```json {"input":{"text":"Caf=E9","action":"decode","charset":"iso-8859-1"}} ``` Result: ```json {"text":"Café","encoding":"iso-8859-1","binary":false,"bytes":4} ``` **Why do I see "=" at the ends of lines?** Quoted-printable lines may be at most 76 characters, so longer lines are split with a trailing "=". Decoding joins them again. **Decoding shows the wrong accented letters. Why?** The email used a different character set. Try "iso-8859-1" or "windows-1252" in place of "utf-8". ### Random generator Make secure random integers, decimals, bytes and dice rolls, or pick from and shuffle a list. - Endpoint: `POST https://setget.net/v1/tools/random` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `random` - Page: https://setget.net/tools/random #### Inputs - `kind` (optional, string, one of `integer`, `float`, `bytes`, `dice`, `pick`, `shuffle`, default `"integer"`): What to make: integer, float, bytes (as hex), dice rolls, a pick from a list, or a shuffled list. - `minimum` (optional, number): Lowest value, included. Defaults to 1 for integers and 0 for floats. - `maximum` (optional, number): Highest value. Included for integers. Defaults to 100 for integers and 1 for floats. - `count` (optional, integer, default `1`, range 1..1000): How many values, rolls or picks to make. - `size` (optional, integer, default `16`, range 1..1024): For bytes: how many bytes in each value. - `notation` (optional, string, default `"1d6"`): For dice: what to roll, such as "d20", "2d6" or "3d6+2". - `items` (optional, array): For pick and shuffle: the list to choose from or reorder. - `unique` (optional, boolean, default `true`): For pick: never choose the same item twice. #### Example 1: Five numbers from 1 to 6 ```json {"input":{"kind":"integer","minimum":1,"maximum":6,"count":5}} ``` #### Example 2: Three six-sided dice plus two ```json {"input":{"kind":"dice","notation":"3d6+2"}} ``` #### Example 3: Two different items from a list ```json {"input":{"kind":"pick","items":["red","green","blue"],"count":2}} ``` #### Example 4: A list in random order ```json {"input":{"kind":"shuffle","items":[1,2,3,4,5]}} ``` #### Example 5: A 256-bit random key as hex ```json {"input":{"kind":"bytes","size":32}} ``` #### Example 6: Three decimals between 0 and 1 ```json {"input":{"kind":"float","count":3}} ``` **How random is it?** Values come from the operating system's cryptographic random source, and are drawn without modulo bias. They are suitable for draws, keys and tokens. **How do I flip a coin?** Use kind "pick" with the items "heads" and "tails", or an integer from 0 to 1. **What dice notation is understood?** NdS with an optional modifier: "2d6" rolls two six-sided dice, "d20" rolls one, and "3d6+2" adds 2 to the total. Up to 100 dice with up to 1,000 sides. ### ROT13 and Caesar cipher Apply ROT13, a Caesar shift of any size, or ROT47 to text. Run it again to decode. - Endpoint: `POST https://setget.net/v1/tools/rot13` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `rot13` - Page: https://setget.net/tools/rot13 #### Inputs - `text` (required, string): The text to shift. - `shift` (optional, integer, default `13`, range -1000..1000): Places to move each letter along the alphabet. 13 is ROT13. Use a negative number to undo a shift. - `rot47` (optional, boolean, default `false`): Use ROT47, which also rotates digits and punctuation. The shift is ignored. #### Example 1 ```json {"input":{"text":"Hello, World!"}} ``` Result: ```json {"text":"Uryyb, Jbeyq!"} ``` #### Example 2: A Caesar shift of 3 ```json {"input":{"text":"Attack at dawn","shift":3}} ``` Result: ```json {"text":"Dwwdfn dw gdzq"} ``` #### Example 3: ROT47 ```json {"input":{"text":"Hello, World!","rot47":true}} ``` Result: ```json {"text":"w6==@[ (@C=5P"} ``` **Is ROT13 secure?** No. It is a puzzle and spoiler-hiding trick, not encryption. Anyone can undo it instantly. **How do I decode?** ROT13 and ROT47 undo themselves: run the result through again. For another shift, use the same number with a minus sign. **What happens to accented letters?** Only the letters A-Z and a-z move. Everything else, including accented letters, is left as it is. ### String bytes and code points Show the UTF-8, UTF-16 and UTF-32 bytes and the Unicode code points of a text. - Endpoint: `POST https://setget.net/v1/tools/string-escape-bytes` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `string-escape-bytes` - Page: https://setget.net/tools/string-escape-bytes #### Inputs - `text` (required, string): The text to inspect. #### Example 1 ```json {"input":{"text":"A€😀"}} ``` Result: ```json {"characters":3,"code_points":["U+0041","U+20AC","U+1F600"],"utf8":"41 e2 82 ac f0 9f 98 80","utf16le":"41 00 ac 20 3d d8 00 de","utf32le":"41 00 00 00 ac 20 00 00 00 f6 01 00","bytes":{"utf8":8,"utf16le":8,"utf32le":12},"ascii":false} ``` #### Example 2: An accent written as a separate combining character ```json {"input":{"text":"é"}} ``` Result: ```json {"characters":2,"code_points":["U+0065","U+0301"],"utf8":"65 cc 81","utf16le":"65 00 01 03","utf32le":"65 00 00 00 01 03 00 00","bytes":{"utf8":3,"utf16le":4,"utf32le":8},"ascii":false} ``` **Why is the byte count larger than the character count?** UTF-8 uses one byte for plain ASCII and two to four for everything else. UTF-16 uses two bytes, or four for characters beyond U+FFFF such as emoji. **Two texts look the same but differ. How can I tell why?** Compare their code points. Common causes are a combining accent in place of a single accented letter, a non-breaking space (U+00A0) or a zero-width character (U+200B). **Which byte order is shown?** Little-endian for UTF-16 and UTF-32, lowest byte first, with no byte order mark. ### Text to binary converter Convert text to binary, octal or decimal byte values and back again. - Endpoint: `POST https://setget.net/v1/tools/binary-text` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `binary-text` - Page: https://setget.net/tools/binary-text #### Inputs - `text` (required, string): The text to convert, or the byte values to turn back into text. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode text to numbers, or decode numbers to text. - `base` (optional, string, one of `binary`, `octal`, `decimal`, default `"binary"`): How each byte is written: binary (8 digits), octal (3 digits) or decimal. #### Example 1 ```json {"input":{"text":"Hi"}} ``` Result: ```json {"text":"01001000 01101001","bytes":2} ``` #### Example 2: Decimal byte values (ASCII codes) ```json {"input":{"text":"Hi","base":"decimal"}} ``` Result: ```json {"text":"72 105","bytes":2} ``` #### Example 3: Binary without spaces ```json {"input":{"text":"0100100001101001","action":"decode"}} ``` Result: ```json {"text":"Hi","encoding":"utf-8","binary":false,"bytes":2} ``` #### Example 4 ```json {"input":{"text":"110 151","action":"decode","base":"octal"}} ``` Result: ```json {"text":"Hi","encoding":"utf-8","binary":false,"bytes":2} ``` **Why does one character give several numbers?** Text is stored as UTF-8 bytes. Plain English letters take one byte each, while accented letters, other scripts and emoji take two to four. **How should the numbers be separated for decoding?** With spaces, commas or line breaks. Binary in full groups of 8 digits and octal in groups of 3 can also be run together. ### ULID generator and decoder Generate ULIDs, sortable 26-character IDs, or decode one to read its creation time. - Endpoint: `POST https://setget.net/v1/tools/ulid` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ulid` - Page: https://setget.net/tools/ulid #### Inputs - `action` (optional, string, one of `generate`, `decode`, default `"generate"`): Generate new ULIDs, or decode an existing one. - `count` (optional, integer, default `1`, range 1..100): How many to generate. - `text` (optional, string): The ULID to decode. - `lowercase` (optional, boolean, default `false`): Write generated ULIDs in lower case. #### Example 1: Three new ULIDs ```json {"input":{"count":3}} ``` #### Example 2: Reading the time inside a ULID ```json {"input":{"action":"decode","text":"01ARYZ6S41TSV4RRFFQ69G5FAV"}} ``` Result: ```json {"ulid":"01ARYZ6S41TSV4RRFFQ69G5FAV","timestamp":"2016-07-30T22:36:16.385Z","unix_ms":1469918176385,"randomness":"d6764c61efb99302bd5b"} ``` **What is inside a ULID?** A 48-bit timestamp in milliseconds followed by 80 random bits, written as 26 Crockford Base32 characters. ULIDs sort by creation time as plain text. **Does a ULID reveal anything?** Yes, its creation time to the millisecond. Anyone can read it, as the decode action shows. **Why are I, L, O and U missing?** Crockford Base32 leaves them out to avoid confusion. When decoding, I and L are read as 1 and O as 0. ### URL encoder and decoder Percent-encode text for use in a URL, or decode it. Works on one value or a whole URL. - Endpoint: `POST https://setget.net/v1/tools/url-encode` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `url-encode` - Page: https://setget.net/tools/url-encode #### Inputs - `text` (required, string): The text or URL to encode or decode. - `action` (optional, string, one of `encode`, `decode`, default `"encode"`): Encode to percent-escapes, or decode them. - `mode` (optional, string, one of `component`, `full`, default `"component"`): "component" treats the text as one value and escapes every reserved character. "full" treats it as a whole URL and keeps : / ? & = # in place. - `plus_for_space` (optional, boolean, default `false`): Treat "+" as a space when decoding, and write spaces as "+" when encoding, as HTML forms do. #### Example 1 ```json {"input":{"text":"hello world & more/stuff?"}} ``` Result: ```json {"text":"hello%20world%20%26%20more%2Fstuff%3F"} ``` #### Example 2: A whole URL keeps its structure ```json {"input":{"text":"https://example.com/café menu?q=a b&lang=en","mode":"full"}} ``` Result: ```json {"text":"https://example.com/caf%C3%A9%20menu?q=a%20b&lang=en"} ``` #### Example 3: Decoding form data ```json {"input":{"text":"q=caf%C3%A9+au+lait%21","action":"decode","plus_for_space":true}} ``` Result: ```json {"text":"q=café au lait!","invalid_utf8":false} ``` **Should I use component or full?** Use "component" for a single value you are putting into a URL, such as a search term. Use "full" when you have a complete URL and only want unsafe characters such as spaces fixed. **When is "+" a space?** Only in form data and query strings written by HTML forms. Elsewhere "+" is a literal plus, so the option is off by default. **Does full mode encode a URL twice?** No. Escapes that are already there, such as %20, are left alone. A stray "%" becomes %25. ### UUID generator Generate UUIDs: random (v4), time-ordered (v7), name-based (v5) or the nil UUID. - Endpoint: `POST https://setget.net/v1/tools/uuid` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `uuid` - Page: https://setget.net/tools/uuid #### Inputs - `version` (optional, string, one of `v4`, `v7`, `v5`, `nil`, default `"v4"`): v4 is random, v7 is time-ordered and sorts by creation time, v5 is always the same for the same namespace and name, nil is all zeros. - `count` (optional, integer, default `1`, range 1..100): How many to generate. v5 and nil always give one. - `name` (optional, string): For v5: the name to turn into a UUID. - `namespace` (optional, string, default `"dns"`): For v5: "dns", "url", "oid", "x500" or any UUID to use as the namespace. - `uppercase` (optional, boolean, default `false`): Write the letters A-F in upper case. - `hyphens` (optional, boolean, default `true`): Keep the hyphens. Turn off for 32 plain hex digits. #### Example 1: Three random UUIDs ```json {"input":{"version":"v4","count":3}} ``` #### Example 2: A name-based UUID is always the same ```json {"input":{"version":"v5","namespace":"dns","name":"python.org"}} ``` Result: ```json {"text":"886313e1-3b8a-5372-9b90-0c9aee199e5d","uuids":["886313e1-3b8a-5372-9b90-0c9aee199e5d"],"version":"v5"} ``` #### Example 3: Time-ordered, compact ```json {"input":{"version":"v7","count":2,"uppercase":true,"hyphens":false}} ``` **Which version should I use?** v4 for general use. v7 when the IDs are database keys, because they sort by creation time. v5 when the same input must always give the same ID. **Can two random UUIDs collide?** In practice, no. A v4 UUID has 122 random bits, so a clash is far less likely than a hardware fault. **Does a v7 UUID reveal anything?** Yes, its creation time to the millisecond. Use v4 if that should stay private. ### UUID inspector Check a UUID and read its version, variant and, for time-based versions, its timestamp. - Endpoint: `POST https://setget.net/v1/tools/uuid-inspect` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `uuid-inspect` - Page: https://setget.net/tools/uuid-inspect #### Inputs - `text` (required, string): The UUID to inspect. Braces, "urn:uuid:" and missing hyphens are accepted. #### Example 1 ```json {"input":{"text":"017f22e2-79b0-7cc3-98c4-dc0c0c07398f"}} ``` Result: ```json {"valid":true,"uuid":"017f22e2-79b0-7cc3-98c4-dc0c0c07398f","version":7,"type":"Unix time-ordered","variant":"RFC 9562","timestamp":"2022-02-22T19:22:22Z","error":null} ``` #### Example 2: Version 1, with braces ```json {"input":{"text":"{C232AB00-9414-11EC-B3C8-9F6BDECED846}"}} ``` Result: ```json {"valid":true,"uuid":"c232ab00-9414-11ec-b3c8-9f6bdeced846","version":1,"type":"time-based","variant":"RFC 9562","timestamp":"2022-02-22T19:22:22Z","error":null} ``` #### Example 3: Not a UUID ```json {"input":{"text":"not-a-uuid"}} ``` Result: ```json {"valid":false,"uuid":null,"version":null,"type":null,"variant":null,"timestamp":null,"error":"A UUID is 32 hex digits, usually written in five groups: 8-4-4-4-12."} ``` **Which UUIDs contain a time?** Versions 1, 6 and 7. Versions 3, 4, 5 and 8 hold no readable timestamp. **What does the variant mean?** It says which layout the UUID follows. Nearly every UUID today is the RFC 9562 variant (formerly RFC 4122). The version number only has meaning for that variant. ## Time, numbers and colour Time zones, dates, cron, units, safe maths and colour conversion. ### Add or subtract time from a date Add or subtract years, months, weeks, days, hours and minutes from a date and get the resulting date. - Endpoint: `POST https://setget.net/v1/tools/date-add` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `date-add` - Page: https://setget.net/tools/date-add #### Inputs - `date` (required, string): The starting date or date and time. - `years` (optional, integer, default `0`, range -9999..9999): Years to add. Negative numbers subtract. - `months` (optional, integer, default `0`, range -120000..120000): Months to add. Negative numbers subtract. - `weeks` (optional, integer, default `0`, range -520000..520000): Weeks to add. Negative numbers subtract. - `days` (optional, integer, default `0`, range -3660000..3660000): Days to add. Negative numbers subtract. - `hours` (optional, integer, default `0`, range -90000000..90000000): Hours to add. Negative numbers subtract. - `minutes` (optional, integer, default `0`, range -90000000..90000000): Minutes to add. Negative numbers subtract. #### Example 1 ```json {"input":{"date":"2026-01-31","months":1}} ``` Result: ```json {"result":"2026-02-28","weekday":"Saturday","clamped":true,"note":"February 2026 has no day 31, so the last day of the month (28) was used."} ``` #### Example 2: Days and hours ```json {"input":{"date":"2026-03-05 09:00","days":10,"hours":-2}} ``` Result: ```json {"result":"2026-03-15T07:00:00","weekday":"Sunday","clamped":false,"note":null} ``` **What happens at the end of a month?** Adding a month to 31 January gives 28 February (29 in a leap year), because February has no 31st. The answer is clamped to the last day of the shorter month, and "clamped" is true. **In what order are the parts added?** Years and months first, then weeks, days, hours and minutes. **Does it follow daylight saving?** No. It works on the clock time as written. Use timezone-convert to move the result between zones. ### Age calculator Work out an exact age in years, months and days on a given date, plus the next birthday. - Endpoint: `POST https://setget.net/v1/tools/age-calculator` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `age-calculator` - Page: https://setget.net/tools/age-calculator #### Inputs - `birth_date` (required, string): Date of birth. - `on` (required, string): The date to work out the age on, usually today. #### Example ```json {"input":{"birth_date":"1990-05-15","on":"2026-03-05"}} ``` Result: ```json {"years":35,"months":9,"days":18,"total_months":429,"total_weeks":1868,"total_days":13078,"born_on":"Tuesday","birthday_today":false,"next_birthday":"2026-05-15","next_birthday_weekday":"Friday","days_until_birthday":71,"turning":36} ``` **How are people born on 29 February handled?** In years without a 29 February the birthday is counted on 28 February. Some countries use 1 March in law, so check local rules if it matters. **Why do I have to send the "on" date?** So the same request always gives the same answer. Send today's date for a current age. ### Aspect ratio calculator Simplify a width and height to a ratio such as 16:9, and scale to a new width or height. - Endpoint: `POST https://setget.net/v1/tools/aspect-ratio` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `aspect-ratio` - Page: https://setget.net/tools/aspect-ratio #### Inputs - `width` (required, number, range ..1000000000.0): The original width. - `height` (required, number, range ..1000000000.0): The original height. - `target_width` (optional, number, range ..1000000000.0): A new width, to get the matching height. - `target_height` (optional, number, range ..1000000000.0): A new height, to get the matching width. #### Example 1 ```json {"input":{"width":1920,"height":1080,"target_width":1280}} ``` Result: ```json {"ratio":"16:9","ratio_width":16,"ratio_height":9,"decimal":1.7778,"orientation":"landscape","nearest_common":"16:9","target":{"width":1280,"height":720}} ``` #### Example 2: Nearly 16:9 ```json {"input":{"width":1366,"height":768}} ``` Result: ```json {"ratio":"683:384","ratio_width":683,"ratio_height":384,"decimal":1.7786,"orientation":"landscape","nearest_common":"16:9","target":null} ``` **Why is 1366 x 768 not exactly 16:9?** Because 1366 / 768 is 1.7786, not 1.7778. The exact ratio is 683:384, and "nearest_common" tells you it is closest to 16:9. **Are the target sizes rounded?** To two decimal places. Round to whole pixels yourself if you need to. ### Business days calculator Count the working days between two dates, or add working days to a date, skipping weekends and holidays. - Endpoint: `POST https://setget.net/v1/tools/business-days` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `business-days` - Page: https://setget.net/tools/business-days #### Inputs - `start` (required, string): The start date. - `end` (optional, string): The end date, to count the working days between the two. - `add` (optional, integer, range -36500..36500): Working days to add to the start date instead. Negative numbers go back. - `weekend` (optional, array, default `["sat", "sun"]`): Days of the week that are not worked. - `holidays` (optional, array): Dates that are not worked, as YYYY-MM-DD. - `inclusive` (optional, boolean, default `true`): When counting, include the end date as well as the start date. #### Example 1 ```json {"input":{"start":"2026-03-02","end":"2026-03-13","holidays":["2026-03-06"]}} ``` Result: ```json {"mode":"count","start":"2026-03-02","end":"2026-03-13","working_days":9,"calendar_days":12,"weekend_days":2,"holidays_skipped":["2026-03-06"]} ``` #### Example 2: Five working days later ```json {"input":{"start":"2026-03-05","add":5}} ``` Result: ```json {"mode":"add","start":"2026-03-05","add":5,"result":"2026-03-12","weekday":"Thursday","calendar_days":7,"holidays_skipped":[]} ``` **Are the start and end dates counted?** Yes, both, when they are working days. This matches NETWORKDAYS in spreadsheets. Set inclusive to false to leave out the end date. **Does it know public holidays?** No. Holidays differ by country and region, so list the dates you want skipped. **My weekend is Friday and Saturday. Can I change it?** Yes. Set weekend to ["fri", "sat"], or to any other days. ### Calculator and expression evaluator Work out an arithmetic expression safely: brackets, powers, functions such as sqrt and log, and variables. - Endpoint: `POST https://setget.net/v1/tools/math-eval` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `math-eval` - Page: https://setget.net/tools/math-eval #### Inputs - `expression` (required, string): The expression to work out. - `variables` (optional, json): Optional named numbers to use in the expression, as an object. #### Example 1 ```json {"input":{"expression":"(2 + 3) * 4 ** 2 / sqrt(16)"}} ``` Result: ```json {"result":20,"text":"20","type":"integer"} ``` #### Example 2: With variables ```json {"input":{"expression":"sqrt(x**2 + y**2)","variables":{"x":3,"y":4}}} ``` Result: ```json {"result":5,"text":"5","type":"integer"} ``` #### Example 3: A very large whole number ```json {"input":{"expression":"2**100"}} ``` Result: ```json {"result":"1267650600228229401496703205376","text":"1267650600228229401496703205376","type":"integer"} ``` #### Example 4: Decimals without noise ```json {"input":{"expression":"0.1 + 0.2"}} ``` Result: ```json {"result":0.3,"text":"0.3","type":"number"} ``` **What does ^ mean?** A power, the same as **. So 2^10 is 1024. **Why is a big answer returned as text?** Whole numbers above 9,007,199,254,740,992 cannot be held exactly by many programs, so "result" is text for those and no digits are lost. **Do the trigonometry functions use degrees or radians?** Radians. Write sin(radians(30)) for 30 degrees. Because pi cannot be stored exactly, sin(pi) gives a tiny number close to 0 instead of exactly 0. **What are the limits?** Expressions up to 1,000 characters, whole numbers up to about 10,000 digits, factorial up to 3,000, and decimal results rounded to 12 significant digits. **Is it safe?** Yes. The expression is parsed and only arithmetic is allowed. No code is ever run. ### Colour contrast checker Check the WCAG contrast ratio between text and background colours, with AA and AAA pass or fail. - Endpoint: `POST https://setget.net/v1/tools/contrast-ratio` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `contrast-ratio` - Page: https://setget.net/tools/contrast-ratio #### Inputs - `foreground` (required, string): The text colour, in any CSS format. - `background` (required, string): The background colour, in any CSS format. #### Example 1 ```json {"input":{"foreground":"#000000","background":"#ffffff"}} ``` Result: ```json {"ratio":21,"text":"21:1","foreground":"#000000","background":"#ffffff","aa":{"normal":true,"large":true},"aaa":{"normal":true,"large":true},"ui_components":true,"suggestion":null} ``` #### Example 2: A grey that just fails ```json {"input":{"foreground":"#777777","background":"#ffffff"}} ``` Result: ```json {"ratio":4.47,"text":"4.47:1","foreground":"#777777","background":"#ffffff","aa":{"normal":false,"large":true},"aaa":{"normal":false,"large":false},"ui_components":true,"suggestion":{"foreground":"#767676","ratio":4.54,"passes":"AA"}} ``` **What ratios are needed?** AA needs 4.5:1 for normal text and 3:1 for large text. AAA needs 7:1 and 4.5:1. Icons and interface components need 3:1. **What counts as large text?** At least 18 point (24 px), or 14 point (about 18.7 px) when bold. **Why does 4.47 not round up to pass?** WCAG compares the exact ratio, so 4.499 fails 4.5. The ratio shown is cut to two decimals, never rounded up. **How is the suggestion chosen?** It keeps the hue and saturation of your text colour and changes the lightness by the smallest amount that reaches AA for normal text. **What about transparent colours?** A see-through foreground is blended over the background first, and a see-through background is blended over white. ### Colour converter Convert a colour between hex, RGB, HSL, HWB, OKLab and OKLCH, and find the nearest CSS colour name. - Endpoint: `POST https://setget.net/v1/tools/color-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `color-convert` - Page: https://setget.net/tools/color-convert #### Inputs - `color` (required, string): A colour in any CSS format: hex, rgb(), hsl(), hwb(), oklab(), oklch() or a name. #### Example 1 ```json {"input":{"color":"#ff0000"}} ``` Result: ```json {"hex":"#ff0000","rgb":"rgb(255, 0, 0)","hsl":"hsl(0, 100%, 50%)","hwb":"hwb(0 0% 0%)","oklab":"oklab(0.628 0.2249 0.1258)","oklch":"oklch(0.628 0.2577 29.23)","alpha":1,"red":255,"green":0,"blue":0,"name":"red","nearest_name":"red","in_gamut":true} ``` #### Example 2: HSL to everything else ```json {"input":{"color":"hsl(270, 50%, 40%)"}} ``` Result: ```json {"hex":"#663399","rgb":"rgb(102, 51, 153)","hsl":"hsl(270, 50%, 40%)","hwb":"hwb(270 20% 40%)","oklab":"oklab(0.4403 0.0882 -0.1339)","oklch":"oklch(0.4403 0.1603 303.37)","alpha":1,"red":102,"green":51,"blue":153,"name":"rebeccapurple","nearest_name":"rebeccapurple","in_gamut":true} ``` **Which colour formats can I type?** Hex (#rgb, #rrggbb, #rrggbbaa), rgb() and rgba(), hsl() and hsla(), hwb(), oklab(), oklch(), and any of the 148 CSS colour names such as "rebeccapurple". **What does "in_gamut" mean?** OKLab and OKLCH can describe colours that an ordinary sRGB screen cannot show. When that happens the colour is clipped to the nearest displayable value and in_gamut is false. **How is the nearest name chosen?** By the smallest distance in OKLab, which matches how different two colours look better than comparing RGB numbers. ### Colour mixer Mix two colours by any ratio in sRGB or OKLab, or lighten or darken a colour by a percentage. - Endpoint: `POST https://setget.net/v1/tools/color-blend` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `color-blend` - Page: https://setget.net/tools/color-blend #### Inputs - `color` (required, string): The starting colour, in any CSS format. - `other` (optional, string): The colour to mix in. Needed for mode "mix". - `mode` (optional, string, one of `mix`, `lighten`, `darken`, default `"mix"`): mix blends towards "other", lighten towards white, darken towards black. - `amount` (optional, number, default `50`, range 0..100): How far to go, in percent. 0 keeps the starting colour, 100 gives the other one. - `space` (optional, string, one of `srgb`, `oklab`, default `"srgb"`): Colour space to mix in. OKLab gives more even, natural steps. #### Example 1 ```json {"input":{"color":"#ff0000","other":"#0000ff"}} ``` Result: ```json {"hex":"#800080","rgb":"rgb(128, 0, 128)","hsl":"hsl(300, 100%, 25%)","alpha":1,"mode":"mix","amount":50,"space":"srgb"} ``` #### Example 2: The same mix in OKLab ```json {"input":{"color":"#ff0000","other":"#0000ff","space":"oklab"}} ``` Result: ```json {"hex":"#8c53a2","rgb":"rgb(140, 83, 162)","hsl":"hsl(283.4, 32.3%, 48.1%)","alpha":1,"mode":"mix","amount":50,"space":"oklab"} ``` #### Example 3: Lighten by 20% ```json {"input":{"color":"#0088cc","mode":"lighten","amount":20}} ``` Result: ```json {"hex":"#33a0d6","rgb":"rgb(51, 160, 214)","hsl":"hsl(200, 66.7%, 52%)","alpha":1,"mode":"lighten","amount":20,"space":"srgb"} ``` **Which colour formats can I type?** Hex (#rgb, #rrggbb, #rrggbbaa), rgb() and rgba(), hsl() and hsla(), hwb(), oklab(), oklch(), and any of the 148 CSS colour names such as "rebeccapurple". **Why do sRGB and OKLab give different mixes?** sRGB averages the raw red, green and blue numbers, which can look dark or muddy in the middle. OKLab is built around how people see colour, so the steps look more even. CSS color-mix() uses OKLab by default. **What do lighten and darken do exactly?** They mix the colour with white or black by the given percentage, which designers call a tint or a shade. ### Colour palette generator Build colour schemes from one base colour: complementary, analogous, triadic and a 50 to 950 shade scale. - Endpoint: `POST https://setget.net/v1/tools/palette-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `palette-generate` - Page: https://setget.net/tools/palette-generate #### Inputs - `color` (required, string): The base colour, in any CSS format. #### Example ```json {"input":{"color":"#0088cc"}} ``` Result: ```json {"base":"#0088cc","complementary":["#0088cc","#cc4400"],"analogous":["#00ccaa","#0088cc","#0022cc"],"triadic":["#0088cc","#cc0088","#88cc00"],"tetradic":["#0088cc","#4400cc","#cc4400","#88cc00"],"square":["#0088cc","#aa00cc","#cc4400","#22cc00"],"split_complementary":["#0088cc","#cc0022","#ccaa00"],"monochromatic":["#004466","#006699","#0088cc","#00aaff","#33bbff"],"scale":{"50":"#ebf5fb","100":"#d6ecf7","200":"#add9ef","300":"#7ac1e4","400":"#3da5d8","500":"#0088cc","600":"#006da3","700":"#00527a","800":"#003652","900":"#001b29","950":"#000e14"}} ``` **Which colour formats can I type?** Hex (#rgb, #rrggbb, #rrggbbaa), rgb() and rgba(), hsl() and hsla(), hwb(), oklab(), oklch(), and any of the 148 CSS colour names such as "rebeccapurple". **How are the schemes built?** By turning the hue on the HSL colour wheel while keeping saturation and lightness: 180 degrees for complementary, 30 either side for analogous, 120 for triadic, 60/180/240 for tetradic, 90 for square and 150/210 for split complementary. **How is the 50 to 950 scale made?** 500 is your base colour. Lighter steps mix in more white and darker steps mix in more black. **What is monochromatic?** The same hue and saturation at five lightness levels: two darker than the base and two lighter. ### Cron expression explainer Explain a cron expression in plain English, for example "At 09:00 on Monday to Friday". - Endpoint: `POST https://setget.net/v1/tools/cron-explain` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `cron-explain` - Page: https://setget.net/tools/cron-explain #### Inputs - `expression` (required, string): A five-field cron expression or a macro such as @daily. #### Example 1 ```json {"input":{"expression":"0 9 * * 1-5"}} ``` Result: ```json {"expression":"0 9 * * 1-5","text":"At 09:00 on Monday to Friday.","fields":{"minute":[0],"hour":[9],"day_of_month":"any","month":"any","day_of_week":[1,2,3,4,5]},"day_rule":"day_of_week","macro":null} ``` #### Example 2: Steps, ranges and names ```json {"input":{"expression":"*/15 9-17 * JAN,JUL MON"}} ``` Result: ```json {"expression":"*/15 9-17 * JAN,JUL MON","text":"Every 15 minutes, during hours 9 to 17 on Monday in January and July.","fields":{"minute":[0,15,30,45],"hour":[9,10,11,12,13,14,15,16,17],"day_of_month":"any","month":[1,7],"day_of_week":[1]},"day_rule":"day_of_week","macro":null} ``` #### Example 3: A macro ```json {"input":{"expression":"@monthly"}} ``` Result: ```json {"expression":"0 0 1 * *","text":"At 00:00 on day 1 of the month.","fields":{"minute":[0],"hour":[0],"day_of_month":[1],"month":"any","day_of_week":"any"},"day_rule":"day_of_month","macro":"@monthly"} ``` **What happens when both day fields are set?** Standard cron runs when either one matches. "0 0 13 * 5" runs on the 13th and also on every Friday, not only on Friday the 13th. The exception is a day field that starts with *, such as */2: then both must match. **Which syntax is supported?** Five fields with values, lists (1,15), ranges (1-5), steps (*/10, 10-40/5), month and weekday names (JAN, MON), and the macros @yearly, @monthly, @weekly, @daily and @hourly. Seconds, years, L, W and # are not supported. ### Cron next run times List the next run times of a cron expression after a given moment, in the time zone you choose. - Endpoint: `POST https://setget.net/v1/tools/cron-next-runs` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `cron-next-runs` - Page: https://setget.net/tools/cron-next-runs #### Inputs - `expression` (required, string): A five-field cron expression or a macro such as @daily. - `start` (required, string): Runs are listed strictly after this date and time. - `zone` (optional, string, default `"UTC"`): Time zone the schedule runs in. - `count` (optional, integer, default `5`, range 1..50): How many run times to list. #### Example 1 ```json {"input":{"expression":"0 9 * * 1-5","start":"2026-03-05T10:00","zone":"Europe/Oslo","count":3}} ``` Result: ```json {"expression":"0 9 * * 1-5","description":"At 09:00 on Monday to Friday.","zone":"Europe/Oslo","after":"2026-03-05T10:00:00+01:00","runs":["2026-03-06T09:00:00+01:00","2026-03-09T09:00:00+01:00","2026-03-10T09:00:00+01:00"],"runs_utc":["2026-03-06T08:00:00Z","2026-03-09T08:00:00Z","2026-03-10T08:00:00Z"],"note":null} ``` #### Example 2: Day of month or day of week ```json {"input":{"expression":"0 0 15 * 1","start":"2026-03-01","count":4}} ``` Result: ```json {"expression":"0 0 15 * 1","description":"At 00:00 on day 15 of the month or on Monday.","zone":"UTC","after":"2026-03-01T00:00:00+00:00","runs":["2026-03-02T00:00:00+00:00","2026-03-09T00:00:00+00:00","2026-03-15T00:00:00+00:00","2026-03-16T00:00:00+00:00"],"runs_utc":["2026-03-02T00:00:00Z","2026-03-09T00:00:00Z","2026-03-15T00:00:00Z","2026-03-16T00:00:00Z"],"note":null} ``` **What happens when both day fields are set?** Standard cron runs when either one matches. "0 0 13 * 5" runs on the 13th and also on every Friday, not only on Friday the 13th. The exception is a day field that starts with *, such as */2: then both must match. **Which syntax is supported?** Five fields with values, lists (1,15), ranges (1-5), steps (*/10, 10-40/5), month and weekday names (JAN, MON), and the macros @yearly, @monthly, @weekly, @daily and @hourly. Seconds, years, L, W and # are not supported. **Why must I send a start time?** So the same request always gives the same answer. Use time-now to get the current time if you want runs from now. **What happens around daylight saving changes?** A run at a clock time that is skipped when clocks go forward does not happen that day. A clock time that occurs twice when clocks go back runs once, the first time. A note is added when a run is skipped. ### CSS gradient generator Build a CSS linear, radial or conic gradient from a list of colour stops, with an SVG preview. - Endpoint: `POST https://setget.net/v1/tools/css-gradient` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `css-gradient` - Page: https://setget.net/tools/css-gradient #### Inputs - `stops` (required, array): Two or more colours in order. Add a position to any of them, as in "#ff0000 25%". - `type` (optional, string, one of `linear`, `radial`, `conic`, default `"linear"`): Kind of gradient. - `angle` (optional, number, range -3600..3600): Angle in degrees. For linear, 0 points up and 90 points right (default 180). For conic, the starting angle (default 0). - `direction` (optional, string): For linear gradients, a direction such as "to right" or "to bottom left" instead of an angle. - `shape` (optional, string, one of `ellipse`, `circle`, default `"ellipse"`): For radial gradients, the shape. #### Example 1 ```json {"input":{"stops":["#ff0000","#0000ff"],"angle":90}} ``` Result: ```json {"css":"linear-gradient(90deg, #ff0000, #0000ff)","declaration":"background: linear-gradient(90deg, #ff0000, #0000ff);","stops":[{"color":"#ff0000","position":0},{"color":"#0000ff","position":100}],"svg":""} ``` #### Example 2: Radial with a positioned stop ```json {"input":{"stops":["gold","tomato 30%","rebeccapurple"],"type":"radial","shape":"circle"}} ``` Result: ```json {"css":"radial-gradient(circle, #ffd700, #ff6347 30%, #663399)","declaration":"background: radial-gradient(circle, #ffd700, #ff6347 30%, #663399);","stops":[{"color":"#ffd700","position":0},{"color":"#ff6347","position":30},{"color":"#663399","position":100}]} ``` **Which colour formats can I type?** Hex (#rgb, #rrggbb, #rrggbbaa), rgb() and rgba(), hsl() and hsla(), hwb(), oklab(), oklch(), and any of the 148 CSS colour names such as "rebeccapurple". **Which way does the angle point?** As in CSS: 0deg runs bottom to top, 90deg left to right, 180deg top to bottom. **Where do stops without a position go?** The first is at 0%, the last at 100%, and the rest are spread evenly between their neighbours, exactly as browsers do. **Why is there no SVG for radial and conic?** The preview is only made for linear gradients. SVG has no conic gradient, and its radial one is sized differently from CSS. ### Current time in any time zone Get the current date and time in one or more time zones, with Unix time and UTC offset. - Endpoint: `POST https://setget.net/v1/tools/time-now` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `time-now` - Page: https://setget.net/tools/time-now #### Inputs - `zones` (optional, array, default `["UTC"]`): IANA time zone names, such as Europe/London. #### Example ```json {"input":{"zones":["UTC","Asia/Tokyo"]}} ``` **Which time zone names work?** IANA names such as Europe/London, America/New_York or Asia/Kolkata. Abbreviations like PST or IST are not accepted because several places share them. ### Date difference calculator Find the time between two dates: total days, weeks, hours, minutes and seconds, plus years, months and days. - Endpoint: `POST https://setget.net/v1/tools/date-diff` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `date-diff` - Page: https://setget.net/tools/date-diff #### Inputs - `start` (required, string): The first date or date and time. - `end` (required, string): The second date or date and time. #### Example 1 ```json {"input":{"start":"2026-01-01","end":"2026-12-25"}} ``` Result: ```json {"start":"2026-01-01","end":"2026-12-25","negative":false,"total":{"days":358,"weeks":51.142857,"hours":8592,"minutes":515520,"seconds":30931200},"calendar":{"years":0,"months":11,"days":24,"hours":0,"minutes":0,"seconds":0},"text":"11 months and 24 days"} ``` #### Example 2: Across a leap year and a month end ```json {"input":{"start":"2024-01-31","end":"2026-03-05"}} ``` Result: ```json {"start":"2024-01-31","end":"2026-03-05","negative":false,"total":{"days":764,"weeks":109.142857,"hours":18336,"minutes":1100160,"seconds":66009600},"calendar":{"years":2,"months":1,"days":5,"hours":0,"minutes":0,"seconds":0},"text":"2 years, 1 month and 5 days"} ``` **Is the end date included?** No. 1 January to 2 January is 1 day. Add one if you want to count both days. **How are months counted?** Whole calendar months are counted from the start date, then the days left over. When the start day does not exist in a later month, the last day of that month is used. **What about time zones?** Times with an offset are compared exactly. Times without one are treated as UTC, so daylight saving changes are not applied. ### Date parser Read a date written in a common human format and return it as ISO 8601 with weekday and week number. - Endpoint: `POST https://setget.net/v1/tools/date-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `date-parse` - Page: https://setget.net/tools/date-parse #### Inputs - `text` (required, string): The date to read, with an optional time. - `day_first` (optional, boolean, default `false`): Read 03/05/2026 as 3 May (day first) instead of 5 March. #### Example 1 ```json {"input":{"text":"March 5, 2026"}} ``` Result: ```json {"iso":"2026-03-05","date":"2026-03-05","time":null,"utc_offset":null,"weekday":"Thursday","day_of_year":64,"iso_week":"2026-W10-4","quarter":1,"leap_year":false,"format":"month day year","note":null} ``` #### Example 2: Day first ```json {"input":{"text":"03/05/2026","day_first":true}} ``` Result: ```json {"iso":"2026-05-03","date":"2026-05-03","time":null,"utc_offset":null,"weekday":"Sunday","day_of_year":123,"iso_week":"2026-W18-7","quarter":2,"leap_year":false,"format":"numeric, day first","note":null} ``` #### Example 3: An email (RFC 2822) date ```json {"input":{"text":"Thu, 05 Mar 2026 14:30:00 +0100"}} ``` Result: ```json {"iso":"2026-03-05T14:30:00+01:00","date":"2026-03-05","time":"14:30:00","utc_offset":"+01:00","weekday":"Thursday","day_of_year":64,"iso_week":"2026-W10-4","quarter":1,"leap_year":false,"format":"day month year","note":null} ``` **Is 03/05/2026 the 5th of March or the 3rd of May?** It depends on the country. By default the month comes first (5 March). Set day_first to true for the day first style used in most of the world (3 May). **What if a number cannot be a month?** In 13/05/2026 the 13 can only be a day, so it is read that way and a note says so. **Are two-digit years accepted?** No. "26" could be 1926 or 2026, so write all four digits. ### Duration parser and formatter Turn a duration such as "1h 30m", "PT1H30M" or "1:30:00" into seconds, ISO 8601 and plain words. - Endpoint: `POST https://setget.net/v1/tools/duration-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `duration-parse` - Page: https://setget.net/tools/duration-parse #### Inputs - `value` (required, string): A duration in words, ISO 8601 or clock form, or a number of seconds to format. #### Example 1 ```json {"input":{"value":"1h 30m"}} ``` Result: ```json {"seconds":5400,"milliseconds":5400000,"iso":"PT1H30M","human":"1 hour and 30 minutes","clock":"1:30:00","parts":{"days":0,"hours":1,"minutes":30,"seconds":0}} ``` #### Example 2: Seconds back to a readable duration ```json {"input":{"value":"187200"}} ``` Result: ```json {"seconds":187200,"milliseconds":187200000,"iso":"P2DT4H","human":"2 days and 4 hours","clock":"52:00:00","parts":{"days":2,"hours":4,"minutes":0,"seconds":0}} ``` **Is 1:30 an hour and a half or a minute and a half?** Two parts are read as minutes:seconds, so 1:30 is 90 seconds. Write 1:30:00 for an hour and a half. **Why are months and years rejected?** They vary in length (28 to 31 days, 365 or 366 days), so they have no exact number of seconds. Use date-add or date-diff when the calendar matters. **How precise is it?** To the millisecond. Smaller fractions are rounded. ### Fraction and decimal converter Convert a decimal to a fraction or a fraction to a decimal, simplify a/b, and show mixed numbers. - Endpoint: `POST https://setget.net/v1/tools/fraction` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `fraction` - Page: https://setget.net/tools/fraction #### Inputs - `value` (required, string): A decimal (0.75), a fraction (6/8), a mixed number (1 3/4) or a repeating decimal (0.1(6)). - `max_denominator` (optional, integer, range 1..1000000000): Largest denominator allowed. The closest fraction within it is returned. #### Example 1 ```json {"input":{"value":"0.75"}} ``` Result: ```json {"fraction":"3/4","numerator":3,"denominator":4,"mixed":"3/4","decimal":0.75,"decimal_text":"0.75","percent":75,"exact":true} ``` #### Example 2: Simplify, with a mixed number ```json {"input":{"value":"14/8"}} ``` Result: ```json {"fraction":"7/4","numerator":7,"denominator":4,"mixed":"1 3/4","decimal":1.75,"decimal_text":"1.75","percent":175,"exact":true} ``` #### Example 3: The closest simple fraction ```json {"input":{"value":"3.14159","max_denominator":10}} ``` Result: ```json {"fraction":"22/7","numerator":22,"denominator":7,"mixed":"3 1/7","decimal":3.14285714286,"decimal_text":"3.(142857)","percent":314.285714286,"exact":false} ``` **What do the brackets in "decimal_text" mean?** The digits in brackets repeat for ever: 0.(3) is 0.333... and 0.1(6) is 0.1666... You can also type a value that way. **When is "exact" false?** When max_denominator forced an approximation. 3.14159 is exactly 314159/100000, but with a limit of 10 the closest fraction is 22/7. ### ISO week number Find the ISO 8601 week number of a date, or the Monday to Sunday date range of a given week. - Endpoint: `POST https://setget.net/v1/tools/iso-week` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `iso-week` - Page: https://setget.net/tools/iso-week #### Inputs - `date` (optional, string): A date to find the week of. Leave empty when sending year and week. - `year` (optional, integer, range 1..9999): ISO year, used together with week. - `week` (optional, integer, range 1..53): ISO week number, used together with year. #### Example 1 ```json {"input":{"date":"2024-12-30"}} ``` Result: ```json {"iso_year":2025,"week":1,"weekday":1,"weekday_name":"Monday","label":"2025-W01","start":"2024-12-30","end":"2025-01-05","weeks_in_year":52} ``` #### Example 2: The dates of a week ```json {"input":{"year":2026,"week":53}} ``` Result: ```json {"iso_year":2026,"week":53,"weekday":null,"weekday_name":null,"label":"2026-W53","start":"2026-12-28","end":"2027-01-03","weeks_in_year":53} ``` **Why is the ISO year different from the calendar year?** ISO weeks run Monday to Sunday and week 1 is the week with the first Thursday of January. So the last days of December can belong to week 1 of the next year, and the first days of January to week 52 or 53 of the previous one. **Which years have 53 weeks?** Years that start on a Thursday, and leap years that start on a Wednesday. ### Number base converter Convert numbers between any bases from 2 to 36, such as binary, octal, decimal and hexadecimal. - Endpoint: `POST https://setget.net/v1/tools/base-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `base-convert` - Page: https://setget.net/tools/base-convert #### Inputs - `value` (required, string): The number to convert. A 0x, 0b or 0o prefix is understood. - `to_base` (required, integer, range 2..36): Base to convert to. - `from_base` (optional, integer, range 2..36): Base the number is written in. Taken from the prefix, or 10, when left empty. - `prefix` (optional, boolean, default `false`): Add 0x, 0b or 0o to the result when the target base is 16, 2 or 8. - `group` (optional, integer, default `0`, range 0..16): Put a space between every this many digits, counted from the right. 0 turns it off. - `uppercase` (optional, boolean, default `false`): Write letter digits in capitals. - `precision` (optional, integer, default `20`, range 0..100): Most digits to give after the point for fractions. #### Example 1 ```json {"input":{"value":"ff","from_base":16,"to_base":2,"group":4}} ``` Result: ```json {"text":"1111 1111","from_base":16,"to_base":2,"decimal":"255","exact":true} ``` #### Example 2: Decimal to hex ```json {"input":{"value":"-255","to_base":16,"prefix":true}} ``` Result: ```json {"text":"-0xff","from_base":10,"to_base":16,"decimal":"-255","exact":true} ``` #### Example 3: A fraction that never ends in binary ```json {"input":{"value":"0.1","to_base":2,"precision":8}} ``` Result: ```json {"text":"0.00011001","from_base":10,"to_base":2,"decimal":"0.1","exact":false} ``` **Why is "exact" false for 0.1 in binary?** Like 1/3 in decimal, 0.1 has no finite form in base 2. The digits are cut off (not rounded) at "precision" places. **How are negative numbers shown?** With a minus sign in front. Two's complement is not used. **Which digits do bases above 10 use?** 0 to 9 and then a to z, so base 16 uses 0-9 and a-f, and base 36 uses 0-9 and a-z. ### Number formatter Format a number with thousands separators, in scientific or compact form, as an ordinal, or in words. - Endpoint: `POST https://setget.net/v1/tools/number-format` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `number-format` - Page: https://setget.net/tools/number-format #### Inputs - `value` (required, string): The number to format. - `style` (optional, string, one of `decimal`, `scientific`, `engineering`, `compact`, `ordinal`, `words`, default `"decimal"`): Output style. - `locale` (optional, string, one of `en`, `de`, `fr`, `ch`, default `"en"`): Separator style for decimal and compact output: en 1,234.5, de 1.234,5, fr 1 234,5, ch 1'234.5. - `decimals` (optional, integer, range 0..20): Digits after the decimal point. Leave empty to keep the number as written. #### Example 1 ```json {"input":{"value":"1234567.891","decimals":2}} ``` Result: ```json {"text":"1,234,567.89","style":"decimal","locale":"en"} ``` #### Example 2: German separators ```json {"input":{"value":"1234567.891","locale":"de","decimals":1}} ``` Result: ```json {"text":"1.234.567,9","style":"decimal","locale":"de"} ``` #### Example 3: Compact ```json {"input":{"value":"1234567","style":"compact"}} ``` Result: ```json {"text":"1.2M","style":"compact","locale":"en"} ``` #### Example 4: Engineering notation ```json {"input":{"value":"0.000047","style":"engineering"}} ``` Result: ```json {"text":"47e-6","style":"engineering","locale":"en"} ``` #### Example 5: In words ```json {"input":{"value":"1234","style":"words"}} ``` Result: ```json {"text":"one thousand two hundred and thirty-four","style":"words","locale":"en"} ``` #### Example 6: Ordinal ```json {"input":{"value":"22","style":"ordinal"}} ``` Result: ```json {"text":"22nd","style":"ordinal","locale":"en"} ``` **How are halves rounded?** Away from zero, the way most people learn at school: 2.5 becomes 3 and 0.125 with two decimals becomes 0.13. **What is engineering notation?** Scientific notation where the exponent is always a multiple of 3, so it lines up with prefixes such as kilo, mega and micro. **Which English is used for words?** British style with "and": one hundred and one. Whole numbers up to one quadrillion (10^15) are supported, and digits after a decimal point are read one by one. ### Percentage calculator Work out percentages: X% of Y, what percent X is of Y, percent change, and increase or decrease by X%. - Endpoint: `POST https://setget.net/v1/tools/percent` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `percent` - Page: https://setget.net/tools/percent #### Inputs - `mode` (required, string, one of `percent_of`, `what_percent`, `change`, `increase`, `decrease`): What to work out: percent_of (X% of Y), what_percent (X is what % of Y), change (from X to Y), increase (Y plus X%) or decrease (Y minus X%). - `x` (required, number, range -1000000000000000.0..1000000000000000.0): X: the percentage, or the first value, depending on the mode. - `y` (required, number, range -1000000000000000.0..1000000000000000.0): Y: the amount the percentage applies to, or the second value. #### Example 1 ```json {"input":{"mode":"percent_of","x":15,"y":80}} ``` Result: ```json {"result":12,"formula":"80 * 15 / 100 = 12","text":"15% of 80 is 12"} ``` #### Example 2: Percent change ```json {"input":{"mode":"change","x":80,"y":100}} ``` Result: ```json {"result":25,"formula":"(100 - 80) / 80 * 100 = 25","text":"From 80 to 100 is an increase of 25%"} ``` #### Example 3: What percent ```json {"input":{"mode":"what_percent","x":30,"y":120}} ``` Result: ```json {"result":25,"formula":"30 / 120 * 100 = 25","text":"30 is 25% of 120"} ``` **Why is a rise from 80 to 100 25%, but a fall from 100 to 80 only 20%?** Percent change is measured against the starting value. 20 is a quarter of 80 but only a fifth of 100. **How do I take 20% off a price?** Use mode "decrease" with x = 20 and y = the price. ### Prime factorisation Split a whole number up to 10^15 into its prime factors, and check whether it is prime. - Endpoint: `POST https://setget.net/v1/tools/prime-factors` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `prime-factors` - Page: https://setget.net/tools/prime-factors #### Inputs - `number` (required, integer, range 1..1000000000000000): A whole number from 1 to 1,000,000,000,000,000. #### Example 1 ```json {"input":{"number":360}} ``` Result: ```json {"number":360,"is_prime":false,"factors":[{"prime":2,"exponent":3},{"prime":3,"exponent":2},{"prime":5,"exponent":1}],"text":"2^3 * 3^2 * 5","divisors":24} ``` #### Example 2: A larger number ```json {"input":{"number":600851475143}} ``` Result: ```json {"number":600851475143,"is_prime":false,"factors":[{"prime":71,"exponent":1},{"prime":839,"exponent":1},{"prime":1471,"exponent":1},{"prime":6857,"exponent":1}],"text":"71 * 839 * 1471 * 6857","divisors":16} ``` #### Example 3: A prime ```json {"input":{"number":1000000007}} ``` Result: ```json {"number":1000000007,"is_prime":true,"factors":[{"prime":1000000007,"exponent":1}],"text":"1000000007","divisors":2} ``` **Is 1 a prime number?** No. A prime has exactly two divisors, 1 and itself, and 1 has only one. It has no prime factors. **What is the largest number accepted?** 1,000,000,000,000,000 (10^15). The answer is exact for every number in that range. **What does "divisors" count?** Every whole number that divides it exactly, including 1 and the number itself. ### Roman numeral converter Convert a number to Roman numerals or Roman numerals to a number, from 1 to 3,999,999. - Endpoint: `POST https://setget.net/v1/tools/roman-numerals` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `roman-numerals` - Page: https://setget.net/tools/roman-numerals #### Inputs - `value` (required, string): A whole number, or a Roman numeral. #### Example 1 ```json {"input":{"value":"2026"}} ``` Result: ```json {"direction":"to_roman","number":2026,"roman":"MMXXVI","overline":"MMXXVI"} ``` #### Example 2: Roman to number ```json {"input":{"value":"MCMXCIV"}} ``` Result: ```json {"direction":"to_number","number":1994,"roman":"MCMXCIV","overline":"MCMXCIV"} ``` #### Example 3: Above 3,999 ```json {"input":{"value":"12345"}} ``` Result: ```json {"direction":"to_roman","number":12345,"roman":"(XII)CCCXLV","overline":"X̅I̅I̅CCCXLV"} ``` **How are numbers above 3,999 written?** With a line over a numeral, which multiplies it by 1,000. "roman" shows the overlined part in brackets, so (XII)CCCXLV is 12,345, and "overline" uses real overline marks, which not every font can display. **Why is IIII rejected?** Only the standard form is accepted, and the standard form of 4 is IV. The error tells you the standard spelling of the number you wrote. **Is there a Roman numeral for zero?** No. The system starts at I (1). ### Statistics calculator Summarise a list of numbers: mean, median, mode, range, variance, standard deviation and quartiles. - Endpoint: `POST https://setget.net/v1/tools/statistics` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `statistics` - Page: https://setget.net/tools/statistics #### Inputs - `numbers` (required, array): The numbers to summarise. #### Example ```json {"input":{"numbers":[2,4,4,4,5,5,7,9]}} ``` Result: ```json {"count":8,"sum":40,"mean":5,"median":4.5,"modes":[4],"min":2,"max":9,"range":7,"variance_population":4,"variance_sample":4.57142857143,"stdev_population":2,"stdev_sample":2.1380899353,"q1":4,"q3":5.5,"iqr":1.5} ``` **Sample or population standard deviation?** Use population when your numbers are the whole group. Use sample when they are a selection from a larger group: it divides by one less than the count, so it comes out a little larger. **How are quartiles worked out?** By linear interpolation between the sorted values, the same method as QUARTILE.INC in spreadsheets. Other methods can give slightly different answers for small lists. **Why is "modes" empty?** When every value appears only once there is no most common value. ### Time zone converter Convert a date and time from one time zone to others, with daylight saving handled for that date. - Endpoint: `POST https://setget.net/v1/tools/timezone-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `timezone-convert` - Page: https://setget.net/tools/timezone-convert #### Inputs - `time` (required, string): The date and time to convert. An offset such as +01:00 or Z overrides from_zone. - `to` (required, array): Time zones to convert to, as IANA names. - `from_zone` (optional, string, default `"UTC"`): Time zone the time is written in. #### Example 1 ```json {"input":{"time":"2026-07-01 12:00","from_zone":"Europe/London","to":["America/New_York","Asia/Tokyo"]}} ``` Result: ```json {"input":{"zone":"Europe/London","iso":"2026-07-01T12:00:00+01:00","utc_offset":"+01:00","status":"ok"},"note":null,"utc":"2026-07-01T11:00:00Z","results":[{"zone":"America/New_York","iso":"2026-07-01T07:00:00-04:00","date":"2026-07-01","time":"07:00:00","weekday":"Wednesday","utc_offset":"-04:00","abbreviation":"EDT","dst":true,"day_shift":0},{"zone":"Asia/Tokyo","iso":"2026-07-01T20:00:00+09:00","date":"2026-07-01","time":"20:00:00","weekday":"Wednesday","utc_offset":"+09:00","abbreviation":"JST","dst":false,"day_shift":0}]} ``` #### Example 2: A time skipped by daylight saving ```json {"input":{"time":"2026-03-29 02:30","from_zone":"Europe/Oslo","to":["UTC"]}} ``` Result: ```json {"input":{"zone":"Europe/Oslo","iso":"2026-03-29T03:30:00+02:00","utc_offset":"+02:00","status":"skipped"},"note":"02:30 on 2026-03-29 does not exist in Europe/Oslo because the clocks go forward. It was read as 03:30 (+02:00).","utc":"2026-03-29T01:30:00Z","results":[{"zone":"UTC","iso":"2026-03-29T01:30:00+00:00","date":"2026-03-29","time":"01:30:00","weekday":"Sunday","utc_offset":"+00:00","abbreviation":"UTC","dst":false,"day_shift":0}]} ``` **Which time zone names work?** IANA names such as Europe/London, America/New_York or Asia/Kolkata. Abbreviations like PST or IST are not accepted because several places share them. **What if the time falls in a daylight saving change?** A time skipped when clocks go forward is moved forward by the size of the change. A time that happens twice when clocks go back uses the first occurrence. Either way "status" and "note" tell you, and you can add an explicit offset to choose. **What does day_shift mean?** It is 1 when the converted time falls on the next calendar day, -1 for the previous day. ### Unit converter Convert between units of length, mass, temperature, area, volume, speed, time, data, pressure, energy and more. - Endpoint: `POST https://setget.net/v1/tools/unit-convert` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `unit-convert` - Page: https://setget.net/tools/unit-convert #### Inputs - `value` (required, number, range -1e+30..1e+30): The amount to convert. - `from_unit` (required, string): Unit to convert from, as a symbol or name. - `to_unit` (required, string): Unit to convert to, as a symbol or name. #### Example 1 ```json {"input":{"value":10,"from_unit":"km","to_unit":"miles"}} ``` Result: ```json {"value":6.21371192237,"text":"10 km = 6.21371192237 mi","from":{"unit":"kilometre","symbol":"km","kind":"length"},"to":{"unit":"mile","symbol":"mi","kind":"length"},"factor":0.621371192237,"formula":"1 km = 0.621371192237 mi","note":null} ``` #### Example 2: Temperature ```json {"input":{"value":100,"from_unit":"C","to_unit":"F"}} ``` Result: ```json {"value":212,"text":"100 °C = 212 °F","from":{"unit":"degree Celsius","symbol":"°C","kind":"temperature"},"to":{"unit":"degree Fahrenheit","symbol":"°F","kind":"temperature"},"factor":null,"formula":"°F = °C * 9 / 5 + 32","note":null} ``` #### Example 3: Binary and decimal data units ```json {"input":{"value":1,"from_unit":"GiB","to_unit":"MB"}} ``` Result: ```json {"value":1073.741824,"text":"1 GiB = 1073.741824 MB","from":{"unit":"gibibyte","symbol":"GiB","kind":"data"},"to":{"unit":"megabyte","symbol":"MB","kind":"data"},"factor":1073.741824,"formula":"1 GiB = 1073.741824 MB","note":null} ``` **What is the difference between MB and MiB?** MB is 1,000,000 bytes (decimal, used for disk sizes and network speeds). MiB is 1,048,576 bytes (binary, used for memory). The same goes for kB and KiB, GB and GiB. **Is a gallon the US or the UK one?** "gal", "pint", "fl oz", "ton" and "mpg" are read as the US measures and a note says so. Use gal_uk, pt_uk, floz_uk, "long ton" and mpg_uk for the imperial ones. **How long are a month and a year?** The average Gregorian lengths: a year is 365.2425 days and a month is one twelfth of that. For real calendar dates use date-add. **How precise are the results?** Conversions use exact definitions and results are rounded to 12 significant digits. ### Unix timestamp converter Convert a Unix timestamp to a date, or a date to a Unix timestamp, in seconds and milliseconds. - Endpoint: `POST https://setget.net/v1/tools/unix-timestamp` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `unix-timestamp` - Page: https://setget.net/tools/unix-timestamp #### Inputs - `value` (required, string): A Unix timestamp, or a date and time to turn into one. - `unit` (optional, string, one of `auto`, `seconds`, `milliseconds`, `microseconds`, `nanoseconds`, default `"auto"`): Unit of a numeric timestamp. "auto" judges by its size. - `zone` (optional, string): Time zone for the "local" output, and for reading a date that has no offset. #### Example 1 ```json {"input":{"value":"1700000000"}} ``` Result: ```json {"direction":"to_date","unit":"seconds","seconds":1700000000,"milliseconds":1700000000000,"utc":"2023-11-14T22:13:20Z","weekday":"Tuesday","local":null} ``` #### Example 2: A date to a timestamp ```json {"input":{"value":"2026-03-05T14:30:00+01:00"}} ``` Result: ```json {"direction":"to_timestamp","unit":null,"seconds":1772717400,"milliseconds":1772717400000,"utc":"2026-03-05T13:30:00Z","weekday":"Thursday","local":null} ``` **How does "auto" tell seconds from milliseconds?** By size. Below 100,000,000,000 is read as seconds, below 100,000,000,000,000 as milliseconds, below 100,000,000,000,000,000 as microseconds, and anything larger as nanoseconds. Set "unit" if your value is unusual. **Which time zone is a date read in?** The offset written in the date wins. Otherwise "zone" is used, and UTC if no zone is given. ## Web and dev Parse URLs and user agents, check versions, patterns, networks and markup. ### .gitignore tester Check which paths a .gitignore file ignores and which line decided each one, using real Git matching rules. - Endpoint: `POST https://setget.net/v1/tools/gitignore-test` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `gitignore-test` - Page: https://setget.net/tools/gitignore-test #### Inputs - `gitignore` (required, string): The contents of the .gitignore file. - `paths` (required, array): Paths relative to the folder holding the .gitignore. End a path with / to mark it as a folder. #### Example 1 ```json {"input":{"gitignore":"# logs and build output\n*.log\n/build/\n!important.log\ndocs/**/*.tmp","paths":["debug.log","important.log","build/app.js","src/build/app.js","docs/a/b/x.tmp"]}} ``` Result: ```json {"results":[{"path":"debug.log","ignored":true,"line":2,"pattern":"*.log"},{"path":"important.log","ignored":false,"line":4,"pattern":"!important.log"},{"path":"build/app.js","ignored":true,"line":3,"pattern":"/build/"},{"path":"src/build/app.js","ignored":false,"line":null,"pattern":null},{"path":"docs/a/b/x.tmp","ignored":true,"line":5,"pattern":"docs/**/*.tmp"}],"ignored":["debug.log","build/app.js","docs/a/b/x.tmp"],"not_ignored":["important.log","src/build/app.js"]} ``` #### Example 2: Ignore everything except one folder (from the Git manual) ```json {"input":{"gitignore":"/*\n!/foo\n/foo/*\n!/foo/bar","paths":["readme.md","foo/baz.txt","foo/bar/x.txt"]}} ``` Result: ```json {"results":[{"path":"readme.md","ignored":true,"line":1,"pattern":"/*"},{"path":"foo/baz.txt","ignored":true,"line":3,"pattern":"/foo/*"},{"path":"foo/bar/x.txt","ignored":false,"line":4,"pattern":"!/foo/bar"}],"ignored":["readme.md","foo/baz.txt"],"not_ignored":["foo/bar/x.txt"]} ``` **Why does my "!" rule not bring a file back?** Git does not look inside an ignored folder, so a file cannot be re-included when one of its parent folders is ignored. Un-ignore the folder first. **When is a pattern tied to the top folder?** When it has a slash at the start or in the middle. A pattern without one, such as *.log, matches at any depth. **How do I say that a path is a folder?** End it with a slash. Patterns that end with a slash match folders only. **Are files already tracked by Git affected?** No. Ignore rules apply only to untracked files. This tool cannot see your repository. ### Cache-Control explainer Explain a Cache-Control header directive by directive and how browsers and CDNs will treat the response. - Endpoint: `POST https://setget.net/v1/tools/cache-control-explain` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `cache-control-explain` - Page: https://setget.net/tools/cache-control-explain #### Inputs - `value` (required, string): The Cache-Control header value (with or without the header name). #### Example 1 ```json {"input":{"value":"public, max-age=3600, s-maxage=86400"}} ``` Result: ```json {"directives":[{"name":"public","value":null,"meaning":"Any cache may store the response, even one that would normally be treated as private."},{"name":"max-age","value":3600,"meaning":"The response is fresh for 1 hour. Caches may reuse it without asking the server during that time."},{"name":"s-maxage","value":86400,"meaning":"Shared caches such as CDNs and proxies may reuse it for 1 day. For them this replaces max-age."}],"summary":"Browsers may reuse it for 1 hour without contacting the server; shared caches such as CDNs may reuse it for 1 day.","warnings":[]} ``` #### Example 2: Conflicting directives ```json {"input":{"value":"Cache-Control: no-store, max-age=60"}} ``` Result: ```json {"directives":[{"name":"no-store","value":null,"meaning":"No cache may store the response at all."},{"name":"max-age","value":60,"meaning":"The response is fresh for 1 minute. Caches may reuse it without asking the server during that time."}],"summary":"Nothing may store this response: browsers and shared caches fetch it from the server every time.","warnings":["no-store wins over the other directives, so max-age and public have no effect."]} ``` **Does no-cache mean "do not cache"?** No. It means "store it, but check with the server before using it". To forbid storing, use no-store. **What is a shared cache?** A cache that serves many people, such as a CDN or a company proxy, as opposed to the private cache in one browser. ### Card number (Luhn) and IBAN validator Check a card number with the Luhn checksum and name its network, or check an IBAN by length and mod-97. - Endpoint: `POST https://setget.net/v1/tools/luhn-iban-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `luhn-iban-validate` - Page: https://setget.net/tools/luhn-iban-validate #### Inputs - `value` (required, string): The card number or IBAN. Spaces and hyphens are ignored. - `kind` (optional, string, one of `auto`, `card`, `iban`, default `"auto"`): What the value is. "auto" decides from its first characters. #### Example 1 ```json {"input":{"value":"GB82 WEST 1234 5698 7654 32"}} ``` Result: ```json {"kind":"iban","valid":true,"country":"GB","check_digits":"82","bban":"WEST12345698765432","length":22,"expected_length":22,"checksum_valid":true,"formatted":"GB82 WEST 1234 5698 7654 32","problems":[]} ``` #### Example 2: A well-known Visa test number ```json {"input":{"value":"4111 1111 1111 1111"}} ``` Result: ```json {"kind":"card","valid":true,"network":"Visa","length":16,"luhn_valid":true,"length_valid":true,"last4":"1111","masked":"************1111","problems":[]} ``` **Does a valid result mean the card or account exists?** No. It only means the number is well formed. Nothing is looked up and no bank or card network is contacted. **Is my card number stored or shown?** The result repeats only the last four digits. Use published test numbers rather than a real card wherever you can. **How is the network found?** From the first digits and the length. Ranges overlap in places, so treat it as a best guess. ### CIDR and subnet calculator Work out the network, netmask, host range and size of an IPv4 or IPv6 CIDR block, and split it into subnets. - Endpoint: `POST https://setget.net/v1/tools/cidr-calc` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `cidr-calc` - Page: https://setget.net/tools/cidr-calc #### Inputs - `cidr` (required, string): The block in CIDR form, such as 192.168.1.0/24 or 2001:db8::/48. - `addresses` (optional, array): IP addresses to test for being inside the block. - `new_prefix` (optional, integer, range 0..128): Split the block into subnets of this prefix length. The first 256 are listed. #### Example 1 ```json {"input":{"cidr":"192.168.1.0/24","addresses":["192.168.1.77","192.168.2.1"]}} ``` Result: ```json {"cidr":"192.168.1.0/24","version":4,"network":"192.168.1.0","prefix":24,"netmask":"255.255.255.0","wildcard":"0.0.0.255","broadcast":"192.168.1.255","first_host":"192.168.1.1","last_host":"192.168.1.254","total_addresses":256,"usable_hosts":254,"scope":"private","is_private":true,"is_loopback":false,"is_link_local":false,"is_reserved":false,"host_bits_set":false,"contains":[{"address":"192.168.1.77","inside":true},{"address":"192.168.2.1","inside":false}],"subnets":null} ``` #### Example 2: Split into /24 subnets ```json {"input":{"cidr":"10.0.0.0/22","new_prefix":24}} ``` Result: ```json {"cidr":"10.0.0.0/22","version":4,"network":"10.0.0.0","prefix":22,"netmask":"255.255.252.0","wildcard":"0.0.3.255","broadcast":"10.0.3.255","first_host":"10.0.0.1","last_host":"10.0.3.254","total_addresses":1024,"usable_hosts":1022,"scope":"private","is_private":true,"is_loopback":false,"is_link_local":false,"is_reserved":false,"host_bits_set":false,"contains":[],"subnets":{"prefix":24,"count":4,"addresses_each":256,"truncated":false,"listed":["10.0.0.0/24","10.0.1.0/24","10.0.2.0/24","10.0.3.0/24"]}} ``` **Why are two addresses not usable in IPv4?** The first address names the network and the last is the broadcast address. A /31 (two hosts, point-to-point links) and a /32 (one host) are the exceptions. **Why are some counts in quotes?** Numbers above 2^53, common for IPv6, are returned as text so that no JSON reader rounds them. **What does scope "mixed" mean?** The block is partly public and partly special-purpose space, for example 0.0.0.0/0. ### Cookie parser Parse Cookie and Set-Cookie headers into JSON, with warnings about missing security attributes. - Endpoint: `POST https://setget.net/v1/tools/cookie-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `cookie-parse` - Page: https://setget.net/tools/cookie-parse #### Inputs - `text` (required, string): One or more Cookie or Set-Cookie headers, one per line. - `kind` (optional, string, one of `auto`, `cookie`, `set-cookie`, default `"auto"`): Which header the lines are. "auto" uses the header name, or guesses from the attributes. #### Example 1 ```json {"input":{"text":"Set-Cookie: sid=abc123; Path=/; HttpOnly; SameSite=None; Max-Age=3600"}} ``` Result: ```json {"count":1,"cookies":[{"source":"set-cookie","name":"sid","value":"abc123","domain":null,"path":"/","expires":null,"max_age":3600,"secure":false,"http_only":true,"same_site":"None","partitioned":false,"session":false,"warnings":["No Secure attribute: the cookie is also sent over unencrypted HTTP.","SameSite=None without Secure: browsers reject this cookie."]}]} ``` #### Example 2: A request header ```json {"input":{"text":"Cookie: theme=dark; sid=abc123"}} ``` Result: ```json {"count":2,"cookies":[{"source":"cookie","name":"theme","value":"dark"},{"source":"cookie","name":"sid","value":"abc123"}]} ``` **What is the difference between Cookie and Set-Cookie?** Set-Cookie is sent by the server and carries one cookie with its attributes. Cookie is sent by the browser and carries only names and values, several per line. **Why is a long Max-Age flagged?** Current browsers shorten any cookie lifetime to about 400 days. **Is an Expires date checked?** It is reported as written. Only Max-Age is used for the lifetime warning, because the result must not depend on the current date. ### CSS formatter Pretty-print minified or messy CSS with one declaration per line and nested @media blocks indented. - Endpoint: `POST https://setget.net/v1/tools/css-format` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `css-format` - Page: https://setget.net/tools/css-format #### Inputs - `text` (required, string): The CSS to format. - `indent` (optional, integer, default `2`, range 1..8): Spaces per indentation level. #### Example ```json {"input":{"text":"a{color:red;margin:0 auto}@media (min-width:600px){a,b{color:blue}}"}} ``` Result: ```json {"text":"a {\n color: red;\n margin: 0 auto;\n}\n\n@media (min-width: 600px) {\n a, b {\n color: blue;\n }\n}"} ``` **Does formatting change what the CSS does?** No. Only white space changes, and a semicolon is added after the last declaration of each block. **Does it handle nesting?** Yes. @media, @supports, @layer, @keyframes and nested rules are indented one level per block. ### CSS minifier Shrink CSS by removing comments and needless white space, leaving strings and url(...) values intact. - Endpoint: `POST https://setget.net/v1/tools/css-minify` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `css-minify` - Page: https://setget.net/tools/css-minify #### Inputs - `text` (required, string): The CSS to minify. #### Example ```json {"input":{"text":"a {\n color: red; /* brand */\n margin: 0 auto;\n}\n\n.b > .c, .d { background: url( img/a b.png ) }"}} ``` Result: ```json {"text":"a{color:red;margin:0 auto}.b > .c,.d{background:url(img/a b.png)}","bytes_before":98,"bytes_after":65,"saved":33,"saved_percent":33.7} ``` **Does it rename or merge anything?** No. Only comments and white space that cannot matter are removed, plus the last semicolon in each block. Selectors, values and their order stay as written. **Which comments survive?** Comments that start with /*! are kept, because that marks licence notices. **Why is the space around > or + kept?** Inside calc() those spaces are required, so they are left alone everywhere to be safe. ### curl command converter Parse a curl command into method, URL, headers, body and options, with equivalent Python and JavaScript code. - Endpoint: `POST https://setget.net/v1/tools/curl-to-json` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `curl-to-json` - Page: https://setget.net/tools/curl-to-json #### Inputs - `command` (required, string): The curl command line. #### Example 1 ```json {"input":{"command":"curl -X POST https://api.example.com/items -H 'Content-Type: application/json' -d '{\"name\":\"x\"}'"}} ``` Result: ```json {"method":"POST","url":"https://api.example.com/items","headers":{"Content-Type":"application/json"},"body":"{\"name\":\"x\"}","form":[],"options":{"follow_redirects":false,"insecure":false,"compressed":false,"auth":null,"timeout":null,"proxy":null},"unknown_flags":[],"python":"import requests\n\nresponse = requests.request(\n 'POST',\n 'https://api.example.com/items',\n headers={\n 'Content-Type': 'application/json',\n },\n data='{\"name\":\"x\"}',\n allow_redirects=False,\n)\nprint(response.status_code)\nprint(response.text)","javascript":"const response = await fetch(\"https://api.example.com/items\", {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\"\n },\n body: \"{\\\"name\\\":\\\"x\\\"}\",\n redirect: \"manual\"\n});\nconsole.log(response.status);\nconsole.log(await response.text());"} ``` #### Example 2: Combined flags, login and -G ```json {"input":{"command":"curl -sL -u sam:secret \"https://example.com/search\" -G -d q=tea --frobnicate"}} ``` Result: ```json {"method":"GET","url":"https://example.com/search?q=tea","headers":{},"body":null,"form":[],"options":{"follow_redirects":true,"insecure":false,"compressed":false,"auth":{"user":"sam","password":"secret"},"timeout":null,"proxy":null},"unknown_flags":["--frobnicate"],"python":"import requests\n\nresponse = requests.request(\n 'GET',\n 'https://example.com/search?q=tea',\n auth=('sam', 'secret'),\n)\nprint(response.status_code)\nprint(response.text)","javascript":"const response = await fetch(\"https://example.com/search?q=tea\", {\n method: \"GET\",\n headers: {\n \"Authorization\": \"Basic c2FtOnNlY3JldA==\"\n }\n});\nconsole.log(response.status);\nconsole.log(await response.text());"} ``` **Is the request sent?** No. The command is only read and translated. Nothing is contacted. **What happens to flags the tool does not know?** They are listed under unknown_flags and otherwise ignored. If such a flag takes a value, that value may be mistaken for the URL, so check the result. **Why does the Python code say allow_redirects=False?** curl does not follow redirects unless you pass -L, while requests does by default. The code keeps the behaviour of your command. **Are files named with @ read?** No. A value such as -d @body.json is kept as written, because this tool cannot see your files. ### Email address syntax checker Check the syntax of an email address, normalise it and flag likely typos, role and disposable addresses. - Endpoint: `POST https://setget.net/v1/tools/email-syntax` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `email-syntax` - Page: https://setget.net/tools/email-syntax #### Inputs - `email` (required, string): The email address to check. #### Example 1 ```json {"input":{"email":"Sam.Jones+news@Gmial.com"}} ``` Result: ```json {"email":"Sam.Jones+news@Gmial.com","valid":true,"problems":[],"normalised":"Sam.Jones+news@gmial.com","local":"Sam.Jones+news","domain":"gmial.com","ascii_domain":"gmial.com","is_idn":false,"suggestion":"Sam.Jones+news@gmail.com","role":false,"disposable":false,"tag":"news","note":"Syntax only. No DNS lookup or mailbox check was made, so the address may still not exist."} ``` #### Example 2: A broken domain ```json {"input":{"email":"info@exa mple..com"}} ``` Result: ```json {"email":"info@exa mple..com","valid":false,"problems":["The domain contains an empty part: two dots in a row, or a dot at the start or end.","The domain part \"exa mple\" may contain only letters, digits and hyphens, and must not start or end with a hyphen."],"normalised":null,"local":"info","domain":"exa mple..com","ascii_domain":null,"is_idn":false,"suggestion":null,"role":true,"disposable":false,"tag":null,"note":"Syntax only. No DNS lookup or mailbox check was made, so the address may still not exist."} ``` **Does "valid" mean the mailbox exists?** No. Only the form of the address is checked. No DNS lookup is made and no mail is sent. **Why is the part before @ not lowercased?** It may be case-sensitive by the standard, though few providers treat it so. Only the domain is safely lowercased. **How are typos found?** The domain is compared with a list of common misspellings and with the names of large providers. It is a suggestion, not an error. **Are quoted names such as "a b"@example.com accepted?** No. They are allowed by the standard but rejected by nearly every mail system, so they are reported as a problem. ### Glob pattern tester Test file paths against a glob pattern with *, ?, **, {a,b} alternatives and [a-z] character classes. - Endpoint: `POST https://setget.net/v1/tools/glob-test` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `glob-test` - Page: https://setget.net/tools/glob-test #### Inputs - `pattern` (required, string): The glob pattern. - `paths` (required, array): The paths to test, with forward slashes. - `case_sensitive` (optional, boolean, default `true`): Treat upper and lower case as different. - `dot` (optional, boolean, default `false`): Let * and ** match names that start with a dot, such as .env. #### Example 1 ```json {"input":{"pattern":"src/**/*.{js,ts}","paths":["src/app.js","src/lib/util.ts","src/lib/readme.md","test/app.js"]}} ``` Result: ```json {"results":[{"path":"src/app.js","matches":true},{"path":"src/lib/util.ts","matches":true},{"path":"src/lib/readme.md","matches":false},{"path":"test/app.js","matches":false}],"matched":["src/app.js","src/lib/util.ts"],"count":2} ``` #### Example 2: A single star stays inside one folder ```json {"input":{"pattern":"*.txt","paths":["notes.txt","docs/notes.txt",".hidden.txt"]}} ``` Result: ```json {"results":[{"path":"notes.txt","matches":true},{"path":"docs/notes.txt","matches":false},{"path":".hidden.txt","matches":false}],"matched":["notes.txt"],"count":1} ``` **What is the difference between * and **?** * matches within one folder name and never crosses a slash. ** as a whole path part matches any number of folders, including none. **Why did a file starting with a dot not match?** As in most shells, wildcards skip hidden names unless the pattern itself starts that part with a dot. Set dot to true to include them. **Does "src/**" match "src" itself?** No. A final ** needs at least one more path part, so it matches everything inside src. ### HTML formatter Re-indent HTML so the nesting is easy to read, leaving pre, textarea, script and style content untouched. - Endpoint: `POST https://setget.net/v1/tools/html-format` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `html-format` - Page: https://setget.net/tools/html-format #### Inputs - `text` (required, string): The HTML to format. - `indent` (optional, integer, default `2`, range 1..8): Spaces per indentation level. #### Example 1 ```json {"input":{"text":"

Hello world

  • One
  • Two
"}} ``` Result: ```json {"text":"
\n

Hello world

\n
    \n
  • One\n
  • Two\n
\n
"} ``` #### Example 2: Preformatted text is not touched ```json {"input":{"text":"
  a\n  b

","indent":4}} ``` Result: ```json {"text":"\n\n \n
  a\n  b
\n
\n \n"} ``` **Does it repair broken HTML?** No. Tags are written back as they were. Missing end tags are not added, and a stray end tag is dropped. **Why are some tags kept on one line?** Inline elements such as a, b and span stay in the flow of the text around them, because moving them to new lines could add visible spaces. **Can the page look different afterwards?** Rarely. Runs of spaces and line breaks between tags are normalised, which only matters where CSS makes white space significant outside pre. ### HTML link extractor List every link and image in HTML, with link text, resolved address, rel values and internal or external. - Endpoint: `POST https://setget.net/v1/tools/html-extract-links` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `html-extract-links` - Page: https://setget.net/tools/html-extract-links #### Inputs - `text` (required, string): The HTML to scan. - `base_url` (optional, string): The address of the page, used to resolve relative links and to tell internal from external. #### Example ```json {"input":{"text":"About us Partner\"The","base_url":"https://example.com/blog/"}} ``` Result: ```json {"links":[{"text":"About us","href":"/about","url":"https://example.com/about","rel":[],"nofollow":false,"kind":"internal"},{"text":"Partner","href":"https://other.example/","url":"https://other.example/","rel":["nofollow","noopener"],"nofollow":true,"kind":"external"}],"images":[{"src":"logo.png","url":"https://example.com/blog/logo.png","alt":null,"missing_alt":true},{"src":"team.jpg","url":"https://example.com/blog/team.jpg","alt":"The team","missing_alt":false}],"counts":{"links":2,"internal":1,"external":1,"images":2,"images_missing_alt":1},"truncated":false} ``` **What are the link kinds?** internal (same site as base_url, ignoring a leading www.), external, anchor (starts with #), other (mailto:, tel:, javascript: and similar) and unknown (an absolute link when no base_url was given). **Is alt="" reported as missing?** No. An empty alt is the correct way to mark a decorative image. Only images with no alt attribute are flagged. **Are the links checked?** No. Nothing is requested, so broken links are not detected. ### HTML minifier Shrink HTML by removing comments and needless white space, without touching pre, textarea, script or style. - Endpoint: `POST https://setget.net/v1/tools/html-minify` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `html-minify` - Page: https://setget.net/tools/html-minify #### Inputs - `text` (required, string): The HTML to minify. - `keep_comments` (optional, boolean, default `false`): Keep ordinary comments. Conditional comments are always kept. #### Example ```json {"input":{"text":"
\n

Hello world

\n \n
 a  b 
\n
"}} ``` Result: ```json {"text":"

Hello world

 a  b 
","bytes_before":80,"bytes_after":53,"saved":27,"saved_percent":33.8} ``` **Is a space between two words ever removed?** No. Runs of white space are collapsed to one space, and removed completely only next to block-level tags such as div or p, where browsers ignore them anyway. **Are inline scripts and styles minified?** No. Their content is copied exactly. Use the CSS minifier for style sheets. ### HTML to plain text Extract readable plain text from HTML: headings, paragraphs, list bullets and table rows, without scripts or menus. - Endpoint: `POST https://setget.net/v1/tools/html-to-text` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `html-to-text` - Page: https://setget.net/tools/html-to-text #### Inputs - `text` (required, string): The HTML to convert. - `links` (optional, boolean, default `false`): Add each link address in brackets after its text. #### Example 1 ```json {"input":{"text":"Demo

Hello & welcome

Read the\n docs.

  • One
  • Two
","links":true}} ``` Result: ```json {"title":"Demo","text":"Hello & welcome\n\nRead the docs (/docs).\n\n- One\n- Two"} ``` #### Example 2: Table rows become tab-separated lines ```json {"input":{"text":"
NameAge
Sam31
"}} ``` Result: ```json {"title":null,"text":"Name\tAge\nSam\t31"} ``` **What is left out?** Everything inside head, script, style, noscript, template, svg, canvas, iframe, object, nav and footer. **Does it fetch the page?** No. Paste the HTML itself. Nothing is downloaded. **How are images handled?** An image is replaced by its alt text, if it has one. ### HTTP headers parser Turn a raw block of HTTP headers into JSON, explain the well-known ones and audit security headers. - Endpoint: `POST https://setget.net/v1/tools/http-headers-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `http-headers-parse` - Page: https://setget.net/tools/http-headers-parse #### Inputs - `text` (required, string): The raw headers, as copied from curl -i or the browser developer tools. - `audit` (optional, boolean, default `false`): List common security headers that are missing. #### Example 1 ```json {"input":{"text":"HTTP/1.1 200 OK\nContent-Type: text/html; charset=UTF-8\nCache-Control: max-age=600"}} ``` Result: ```json {"start_line":{"type":"response","version":"1.1","status":200,"reason":"OK"},"count":2,"headers":{"content-type":"text/html; charset=UTF-8","cache-control":"max-age=600"},"explained":{"content-type":{"meaning":"The format of the body, and for text its character set.","media_type":"text/html","charset":"utf-8","boundary":null},"cache-control":{"meaning":"How browsers and shared caches may store and reuse the response.","directives":[{"name":"max-age","value":600,"meaning":"The response is fresh for 10 minutes. Caches may reuse it without asking the server during that time."}],"summary":"Browsers may reuse it for 10 minutes without contacting the server; shared caches such as CDNs may reuse it for 10 minutes.","warnings":[]}},"skipped":[],"audit":null} ``` #### Example 2: Security header audit ```json {"input":{"text":"x-frame-options: DENY\nx-content-type-options: nosniff\nreferrer-policy: no-referrer","audit":true}} ``` Result: ```json {"start_line":null,"count":3,"headers":{"x-frame-options":"DENY","x-content-type-options":"nosniff","referrer-policy":"no-referrer"},"explained":{"x-frame-options":{"meaning":"Whether other sites may show this page inside a frame."},"x-content-type-options":{"meaning":"With \"nosniff\", browsers do not guess a different content type."},"referrer-policy":{"meaning":"How much of the page address is sent to other sites in the Referer header."}},"skipped":[],"audit":{"present":["X-Content-Type-Options","X-Frame-Options","Referrer-Policy"],"missing":[{"header":"Strict-Transport-Security","why":"Tells browsers to use HTTPS only. Without it a visit can be downgraded to plain HTTP."},{"header":"Content-Security-Policy","why":"Limits where scripts and other resources may load from, which contains injected code."},{"header":"Permissions-Policy","why":"Switches off browser features the page does not need, such as camera or location."}]}} ``` **What happens to a header that appears twice?** Its values are returned as a list, in the order they appeared. **Does the audit prove a site is secure?** No. It only reports which widely recommended headers are absent from the text you pasted. **Is the body read?** No. Parsing stops at the first blank line after the headers. ### HTTP request code builder Turn a method, URL, headers and body into ready-to-run curl, Python requests and JavaScript fetch code. - Endpoint: `POST https://setget.net/v1/tools/http-request-build` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `http-request-build` - Page: https://setget.net/tools/http-request-build #### Inputs - `url` (required, string): The address to request. - `method` (optional, string, default `"GET"`): The HTTP method. - `headers` (optional, json): Request headers as a JSON object of names and values. - `body` (optional, string): The request body, exactly as it should be sent. #### Example 1 ```json {"input":{"url":"https://api.example.com/items","method":"post","headers":{"Content-Type":"application/json"},"body":"{\"name\": \"Sam's\"}"}} ``` Result: ```json {"curl":"curl -X POST 'https://api.example.com/items' \\\n -H 'Content-Type: application/json' \\\n --data-raw '{\"name\": \"Sam'\\''s\"}'","python":"import requests\n\nresponse = requests.request(\n 'POST',\n 'https://api.example.com/items',\n headers={\n 'Content-Type': 'application/json',\n },\n data='{\"name\": \"Sam\\'s\"}',\n)\nprint(response.status_code)\nprint(response.text)","javascript":"const response = await fetch(\"https://api.example.com/items\", {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\"\n },\n body: \"{\\\"name\\\": \\\"Sam's\\\"}\"\n});\nconsole.log(response.status);\nconsole.log(await response.text());"} ``` #### Example 2: A plain GET ```json {"input":{"url":"https://example.com/health"}} ``` Result: ```json {"curl":"curl 'https://example.com/health'","python":"import requests\n\nresponse = requests.request(\n 'GET',\n 'https://example.com/health',\n)\nprint(response.status_code)\nprint(response.text)","javascript":"const response = await fetch(\"https://example.com/health\", {\n method: \"GET\"\n});\nconsole.log(response.status);\nconsole.log(await response.text());"} ``` **Is the request sent?** No. Only the code is produced. Run it yourself to make the request. **How are quotes in my data handled?** Each snippet escapes them for its own language, so the body arrives exactly as you wrote it. **Why is my JSON body a string?** The body is sent as written, byte for byte. Set the Content-Type header yourself, for example to application/json. ### HTTP status code lookup Look up an HTTP status code or search by words: meaning, typical cause and whether to retry. - Endpoint: `POST https://setget.net/v1/tools/http-status` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `http-status` - Page: https://setget.net/tools/http-status #### Inputs - `query` (required, string): A status code such as 404, or words to search for such as "too many requests". #### Example 1 ```json {"input":{"query":"429"}} ``` Result: ```json {"query":"429","count":1,"matches":[{"code":429,"name":"Too Many Requests","class":"client error","official":true,"meaning":"The client has sent too many requests in a given time.","cause":"A rate limit was hit; the Retry-After header often says how long to wait.","retry":true}]} ``` #### Example 2: Search by a word ```json {"input":{"query":"teapot"}} ``` Result: ```json {"query":"teapot","count":1,"matches":[{"code":418,"name":"I'm a teapot","class":"client error","official":false,"meaning":"A joke code from an April Fools specification; some sites use it to refuse automated requests.","cause":"An Easter egg, or a server that blocks bots with it.","retry":false}]} ``` **What does "official" mean?** Official codes are in the IANA registry. The others are in real use by nginx, Cloudflare and some frameworks but are not standardised. **What does "retry" tell me?** Whether sending the same request again, after a pause, has a fair chance of working. For most 4xx codes you must change the request first. ### IP address inspector Classify an IPv4 or IPv6 address (private, loopback, CGNAT and more) and show its integer, hex and reverse-DNS forms. - Endpoint: `POST https://setget.net/v1/tools/ip-inspect` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ip-inspect` - Page: https://setget.net/tools/ip-inspect #### Inputs - `ip` (required, string): The IP address to inspect. #### Example 1 ```json {"input":{"ip":"192.168.1.10"}} ``` Result: ```json {"ip":"192.168.1.10","version":4,"type":"private","is_private":true,"is_loopback":false,"is_link_local":false,"is_multicast":false,"is_cgnat":false,"is_documentation":false,"is_reserved":false,"is_public":false,"integer":3232235786,"hex":"0xC0A8010A","binary":"11000000.10101000.00000001.00001010","exploded":"192.168.1.10","reverse_dns":"10.1.168.192.in-addr.arpa","ipv4_mapped":"::ffff:192.168.1.10","embedded_ipv4":null} ``` #### Example 2: An IPv6 documentation address ```json {"input":{"ip":"2001:DB8::1"}} ``` Result: ```json {"ip":"2001:db8::1","version":6,"type":"documentation","is_private":false,"is_loopback":false,"is_link_local":false,"is_multicast":false,"is_cgnat":false,"is_documentation":true,"is_reserved":false,"is_public":false,"integer":"42540766411282592856903984951653826561","hex":"0x20010DB8000000000000000000000001","binary":null,"exploded":"2001:0db8:0000:0000:0000:0000:0000:0001","reverse_dns":"1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.arpa","ipv4_mapped":null,"embedded_ipv4":null} ``` **Is anything looked up?** No. The reverse-DNS value is only the name you would query. No DNS request is made and no location or owner is found. **What is CGNAT?** The range 100.64.0.0/10, which internet providers use between their own equipment and customers. It is not reachable from the internet. **Why is the IPv6 integer in quotes?** It has up to 39 digits, far more than most JSON readers can hold exactly. ### JSON-LD structured data checker Check JSON-LD structured data: valid JSON, @context, @type and the usual properties of common schema.org types. - Endpoint: `POST https://setget.net/v1/tools/jsonld-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `jsonld-validate` - Page: https://setget.net/tools/jsonld-validate #### Inputs - `text` (required, string): JSON-LD, or HTML that contains application/ld+json script blocks. #### Example 1 ```json {"input":{"text":"{\"@context\":\"https://schema.org\",\"@type\":\"Organization\",\"name\":\"Example\"}"}} ``` Result: ```json {"valid":false,"blocks":1,"items":[{"block":1,"type":"Organization","known_type":true,"problems":["Organization needs the property \"url\"."],"missing_recommended":["logo","sameAs","contactPoint"]}],"problems":[],"note":"This is a structural check only. It is not Google's Rich Results Test and does not guarantee rich results."} ``` #### Example 2: Taken from a script block in HTML ```json {"input":{"text":""}} ``` Result: ```json {"valid":true,"blocks":1,"items":[{"block":1,"type":"FAQPage","known_type":true,"problems":[],"missing_recommended":[]}],"problems":[],"note":"This is a structural check only. It is not Google's Rich Results Test and does not guarantee rich results."} ``` **Is this the same as the Rich Results Test?** No. It checks structure: that the JSON parses, that @context and @type are present, and that common types carry their usual properties. Search engines apply further rules of their own. **Which types are checked in detail?** Article (also NewsArticle and BlogPosting), Product, FAQPage, BreadcrumbList, Organization, SoftwareApplication and WebSite. Other types are only checked for @context and @type. **Does a missing recommended property make it invalid?** No. Only problems do. Recommended properties are listed so you can add them. ### Meta tag generator Generate ready-to-paste HTML meta tags: title, description, canonical, Open Graph and Twitter card. - Endpoint: `POST https://setget.net/v1/tools/meta-tags-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `meta-tags-generate` - Page: https://setget.net/tools/meta-tags-generate #### Inputs - `title` (required, string): The page title. - `description` (optional, string): A one or two sentence summary of the page. - `url` (optional, string): The canonical address of the page. - `image` (optional, string): The address of the preview image (1200 x 630 pixels works well). - `site_name` (optional, string): The name of the site. - `type` (optional, string, one of `website`, `article`, `product`, `profile`, `book`, `video.other`, `music.song`, default `"website"`): The Open Graph type. - `twitter_card` (optional, string, one of `auto`, `summary`, `summary_large_image`, default `"auto"`): The Twitter card style. "auto" picks the large image card when there is an image. - `twitter_site` (optional, string): The @name of the site on X (Twitter). #### Example ```json {"input":{"title":"Tools & more","description":"Small web utilities.","url":"https://example.com/tools"}} ``` Result: ```json {"text":"Tools & more\n\n\n\n\n\n\n\n\n","warnings":["There is no image. Link previews look much better with one."]} ``` **Where do the tags go?** Inside the head element of the page. **Why the limits of 60 and 160 characters?** Search results usually cut titles at about 60 characters and descriptions at about 160. Longer text still works but may be shortened. **Must the image address be absolute?** Yes. Preview generators do not resolve relative addresses, so start it with https://. ### MIME type lookup Find the MIME type for a file extension, or the extensions for a MIME type, and whether it compresses well. - Endpoint: `POST https://setget.net/v1/tools/mime-lookup` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `mime-lookup` - Page: https://setget.net/tools/mime-lookup #### Inputs - `query` (required, string): A file extension, a file name or a MIME type. #### Example 1 ```json {"input":{"query":"photo.webp"}} ``` Result: ```json {"query":"photo.webp","direction":"extension","extension":"webp","mime":"image/webp","extensions":["webp"],"kind":"binary","compressible":false} ``` #### Example 2: From type to extensions ```json {"input":{"query":"application/json"}} ``` Result: ```json {"query":"application/json","direction":"mime","extension":"json","mime":"application/json","extensions":["json","map"],"kind":"text","compressible":true} ``` **What does "compressible" mean?** Whether gzip or Brotli usually makes the file noticeably smaller. Text formats do; images, video, archives and WOFF fonts are already compressed. **Is the file inspected?** No. The answer comes only from the name or type you enter. ### Open Graph and meta tag extractor Pull the title, description, canonical link, Open Graph, Twitter card, favicons and JSON-LD out of HTML. - Endpoint: `POST https://setget.net/v1/tools/og-tags-extract` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `og-tags-extract` - Page: https://setget.net/tools/og-tags-extract #### Inputs - `text` (required, string): The HTML of the page (the head section is enough). #### Example ```json {"input":{"text":"Demo"}} ``` Result: ```json {"title":"Demo","description":"A demo page.","canonical":"https://example.com/","og":{"og:title":"Demo page","og:image":"https://example.com/a.png"},"twitter":{"twitter:card":"summary"},"favicons":[{"rel":"icon","href":"/favicon.ico","sizes":null,"type":null}],"json_ld":[{"@context":"https://schema.org","@type":"WebSite","name":"Demo"}],"missing":["og:description","og:url","og:type"]} ``` **What does "missing" list?** Tags that link previews in chat apps and social networks rely on: title, description, canonical, og:title, og:description, og:image, og:url, og:type and twitter:card. **What if a tag appears twice?** The first value is used, as most preview generators do. **What happens to broken JSON-LD?** The block is listed as {"error": ...} so you can see that it exists but does not parse. ### robots.txt tester Check which URL paths a robots.txt file allows for a crawler, and which rule decided, following RFC 9309. - Endpoint: `POST https://setget.net/v1/tools/robots-test` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `robots-test` - Page: https://setget.net/tools/robots-test #### Inputs - `robots` (required, string): The contents of the robots.txt file. - `user_agent` (required, string): The crawler to test as: a product name such as Googlebot, or a full User-Agent string. - `paths` (required, array): URL paths (or full URLs) to test. #### Example 1: The example from RFC 9309, as foobot ```json {"input":{"robots":"User-agent: *\nDisallow: *.gif$\nDisallow: /example/\nAllow: /publications/\n\nUser-agent: foobot\nDisallow: /\nAllow: /example/page.html\nAllow: /example/allowed.gif\n\nSitemap: https://example.com/sitemap.xml","user_agent":"foobot","paths":["/example/page.html","/example/allowed.gif","/","/publications/"]}} ``` Result: ```json {"user_agent":"foobot","matched_group":"foobot","results":[{"path":"/example/page.html","allowed":true,"rule":"Allow: /example/page.html","line":8},{"path":"/example/allowed.gif","allowed":true,"rule":"Allow: /example/allowed.gif","line":9},{"path":"/","allowed":false,"rule":"Disallow: /","line":7},{"path":"/publications/","allowed":false,"rule":"Disallow: /","line":7}],"crawl_delay":null,"sitemaps":["https://example.com/sitemap.xml"]} ``` #### Example 2: Wildcards for any other crawler ```json {"input":{"robots":"User-agent: *\nDisallow: *.gif$\nDisallow: /example/\nAllow: /publications/","user_agent":"Mozilla/5.0 (compatible; OtherBot/1.0)","paths":["/example/page.html","/images/logo.gif","/images/logo.gif?v=2","/about"]}} ``` Result: ```json {"user_agent":"Mozilla/5.0 (compatible; OtherBot/1.0)","matched_group":"*","results":[{"path":"/example/page.html","allowed":false,"rule":"Disallow: /example/","line":3},{"path":"/images/logo.gif","allowed":false,"rule":"Disallow: *.gif$","line":2},{"path":"/images/logo.gif?v=2","allowed":true,"rule":null,"line":null},{"path":"/about","allowed":true,"rule":null,"line":null}],"crawl_delay":null,"sitemaps":[]} ``` **Which rule wins when several match?** The one with the longest path pattern. If an Allow and a Disallow are equally long, Allow wins. **Which group applies to my crawler?** The group whose User-agent name appears in the name you test, the longest such name if there are several. Only if none matches does the * group apply. Groups are never combined with *. **Is the file fetched from my site?** No. Paste its contents. Nothing is downloaded. **Does robots.txt block access?** No. It is a request that well-behaved crawlers follow. It is not access control. ### SemVer compare and sort Compare two semantic versions, including prerelease precedence, or sort a list of versions. - Endpoint: `POST https://setget.net/v1/tools/semver-compare` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `semver-compare` - Page: https://setget.net/tools/semver-compare #### Inputs - `a` (optional, string): The first version. - `b` (optional, string): The second version. - `versions` (optional, array): A list of versions to sort from lowest to highest. - `descending` (optional, boolean, default `false`): Sort from highest to lowest. #### Example 1 ```json {"input":{"a":"1.0.0-alpha","b":"1.0.0"}} ``` Result: ```json {"comparison":{"a":{"version":"1.0.0-alpha","major":1,"minor":0,"patch":0,"prerelease":["alpha"],"build":[]},"b":{"version":"1.0.0","major":1,"minor":0,"patch":0,"prerelease":[],"build":[]},"result":-1,"relation":"a < b","difference":"prerelease"},"sorted":null} ``` #### Example 2: The precedence example from the SemVer specification ```json {"input":{"versions":["1.0.0","1.0.0-rc.1","1.0.0-beta.11","1.0.0-beta.2","1.0.0-beta","1.0.0-alpha.beta","1.0.0-alpha.1","1.0.0-alpha"]}} ``` Result: ```json {"comparison":null,"sorted":["1.0.0-alpha","1.0.0-alpha.1","1.0.0-alpha.beta","1.0.0-beta","1.0.0-beta.2","1.0.0-beta.11","1.0.0-rc.1","1.0.0"]} ``` **Why is 1.0.0-beta.11 higher than 1.0.0-beta.2?** Numeric prerelease parts are compared as numbers, not as text. **Does build metadata matter?** No. Anything after "+" is ignored for precedence, so 1.0.0+a and 1.0.0+b are equal. The difference is then reported as "build". **Is a leading "v" allowed?** Yes, v1.2.3 is read as 1.2.3. Everything else must follow SemVer 2.0.0 strictly. ### SemVer range tester Test versions against an npm-style range such as ^1.2.3 or >=1.0 <2 and find the highest match. - Endpoint: `POST https://setget.net/v1/tools/semver-range` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `semver-range` - Page: https://setget.net/tools/semver-range #### Inputs - `range` (required, string): The range, in npm syntax: ^, ~, x-ranges, hyphen ranges, comparators and ||. - `versions` (required, array): The versions to test. - `include_prerelease` (optional, boolean, default `false`): Let prerelease versions match like ordinary versions. #### Example 1 ```json {"input":{"range":"^1.2.3","versions":["1.2.2","1.2.3","1.9.0","2.0.0","1.3.0-beta.1"]}} ``` Result: ```json {"range":"^1.2.3","normalised":">=1.2.3 <2.0.0-0","matching":["1.2.3","1.9.0"],"not_matching":["1.2.2","2.0.0","1.3.0-beta.1"],"invalid":[],"highest":"1.9.0"} ``` #### Example 2: Tilde, hyphen range and "or" ```json {"input":{"range":"~1.2 || 2.0.0 - 2.5","versions":["1.2.9","1.3.0","2.5.7","2.6.0"]}} ``` Result: ```json {"range":"~1.2 || 2.0.0 - 2.5","normalised":">=1.2.0 <1.3.0-0 || >=2.0.0 <2.6.0-0","matching":["1.2.9","2.5.7"],"not_matching":["1.3.0","2.6.0"],"invalid":[],"highest":"2.5.7"} ``` **What is the difference between ^ and ~?** ^1.2.3 allows changes that keep the first non-zero number (up to, not including, 2.0.0). ~1.2.3 allows only patch updates (below 1.3.0). **Why does a prerelease not match?** As in npm, 1.3.0-beta.1 matches only if the range itself names a prerelease of 1.3.0. Turn on include_prerelease to change that. **What does "<2.0.0-0" mean?** Below every 2.0.0 version, including its prereleases. It is how npm writes an exclusive upper bound. ### Sitemap validator Validate sitemap XML: structure, absolute URLs, dates, changefreq, priority, duplicates and the 50,000 URL limit. - Endpoint: `POST https://setget.net/v1/tools/sitemap-validate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `sitemap-validate` - Page: https://setget.net/tools/sitemap-validate #### Inputs - `text` (required, string): The sitemap XML (a urlset or a sitemapindex). #### Example 1 ```json {"input":{"text":"\n\n https://example.com/2026-01-150.8\n /aboutsometimes\n"}} ``` Result: ```json {"valid":false,"type":"urlset","entries":2,"with_lastmod":1,"problem_count":2,"problems":[{"line":4,"message":"The address \"/about\" is not absolute. It must start with http:// or https://."},{"line":4,"message":"changefreq \"sometimes\" is not allowed. Use always, hourly, daily, weekly, monthly, yearly or never."}]} ``` #### Example 2: A sitemap index ```json {"input":{"text":"https://example.com/sitemap-1.xml"}} ``` Result: ```json {"valid":true,"type":"sitemapindex","entries":1,"with_lastmod":0,"problem_count":0,"problems":[]} ``` **Are the listed pages checked?** No. Only the XML you paste is examined. No address is requested. **Why is a DOCTYPE rejected?** Sitemaps do not need one, and DOCTYPE and entity declarations are a known way to attack XML readers, so this tool refuses them. **Which date formats are accepted for lastmod?** The W3C formats: 2026-01-15, 2026-01, 2026, or a full time such as 2026-01-15T09:30:00+00:00. **How many problems are listed?** The first 100. problem_count gives the total. ### URL joiner Resolve relative links such as ../page or ?q=1 against a base URL, following RFC 3986. - Endpoint: `POST https://setget.net/v1/tools/url-join` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `url-join` - Page: https://setget.net/tools/url-join #### Inputs - `base` (required, string): The absolute URL the links are relative to. - `references` (required, array): Relative or absolute links to resolve. #### Example 1: The examples from RFC 3986 ```json {"input":{"base":"http://a/b/c/d;p?q","references":["g","../g","/g","//g","?y","#s","../../../g"]}} ``` Result: ```json {"base":"http://a/b/c/d;p?q","results":[{"reference":"g","url":"http://a/b/c/g"},{"reference":"../g","url":"http://a/b/g"},{"reference":"/g","url":"http://a/g"},{"reference":"//g","url":"http://g"},{"reference":"?y","url":"http://a/b/c/d;p?y"},{"reference":"#s","url":"http://a/b/c/d;p?q#s"},{"reference":"../../../g","url":"http://a/g"}]} ``` #### Example 2 ```json {"input":{"base":"https://example.com/docs/guide/","references":["intro.html","../img/logo.png"]}} ``` Result: ```json {"base":"https://example.com/docs/guide/","results":[{"reference":"intro.html","url":"https://example.com/docs/guide/intro.html"},{"reference":"../img/logo.png","url":"https://example.com/docs/img/logo.png"}]} ``` **Why does the last part of the base disappear?** A relative link replaces everything after the last slash of the base. End the base with a slash if its last part is a folder. ### URL normaliser Clean a URL into a standard form: lowercase host, no default port, resolved dots, optional tracking removal. - Endpoint: `POST https://setget.net/v1/tools/url-normalize` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `url-normalize` - Page: https://setget.net/tools/url-normalize #### Inputs - `url` (required, string): The URL to normalise. - `strip_tracking` (optional, boolean, default `false`): Remove tracking parameters such as utm_source, fbclid and gclid. - `sort_query` (optional, boolean, default `false`): Sort query parameters by name. - `drop_fragment` (optional, boolean, default `false`): Remove the #fragment at the end. - `remove_trailing_slash` (optional, boolean, default `false`): Remove a slash at the end of the path (the root "/" is kept). #### Example 1 ```json {"input":{"url":"HTTPS://Example.COM:443/a/./b/../c?b=2&a=1#top","sort_query":true,"drop_fragment":true}} ``` Result: ```json {"url":"https://example.com/a/c?a=1&b=2","changed":true,"changes":["Lowercased the scheme.","Lowercased the host.","Removed the default port 443.","Resolved the dot segments in the path.","Sorted the query parameters.","Removed the fragment."]} ``` #### Example 2: Strip tracking parameters ```json {"input":{"url":"https://example.com/page/?utm_source=news&id=7&fbclid=abc","strip_tracking":true,"remove_trailing_slash":true}} ``` Result: ```json {"url":"https://example.com/page?id=7","changed":true,"changes":["Removed tracking parameters: utm_source, fbclid.","Removed the trailing slash."]} ``` **Which parameters count as tracking?** Anything starting with utm_, plus well-known click identifiers such as fbclid, gclid, msclkid, mc_eid and igshid. **Could two normalised URLs still be the same page?** Yes. Only changes that are safe by the URL standard are made. A site may still serve the same page at several addresses. ### URL parser Split a URL into scheme, login, host, port, path, query parameters and fragment. - Endpoint: `POST https://setget.net/v1/tools/url-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `url-parse` - Page: https://setget.net/tools/url-parse #### Inputs - `url` (required, string): The URL to take apart. #### Example 1 ```json {"input":{"url":"https://user:pw@Example.com:8443/a/b%20c?x=1&x=2&y=#top"}} ``` Result: ```json {"scheme":"https","username":"user","password":"pw","host":"example.com","host_type":"domain","is_idn":false,"ascii_host":"example.com","unicode_host":"example.com","port":8443,"default_port":443,"path":"/a/b%20c","segments":["a","b c"],"query":"x=1&x=2&y=","params":{"x":["1","2"],"y":""},"fragment":"top","origin":"https://example.com:8443"} ``` #### Example 2: An internationalised domain name ```json {"input":{"url":"http://bücher.example/"}} ``` Result: ```json {"scheme":"http","username":null,"password":null,"host":"bücher.example","host_type":"domain","is_idn":true,"ascii_host":"xn--bcher-kva.example","unicode_host":"bücher.example","port":null,"default_port":80,"path":"/","segments":[],"query":null,"params":{},"fragment":null,"origin":"http://xn--bcher-kva.example"} ``` **Is the URL fetched?** No. The text is only split into its parts. Nothing is requested or looked up. **Why is a parameter sometimes a list?** A name that appears more than once in the query becomes a list of its values, in order. **What is the origin?** Scheme, host and port together. Browsers use it to decide which pages may share data. ### User agent parser Read a User-Agent string: browser, engine, operating system, device type and known bots or AI agents. - Endpoint: `POST https://setget.net/v1/tools/user-agent-parse` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `user-agent-parse` - Page: https://setget.net/tools/user-agent-parse #### Inputs - `user_agent` (required, string): The User-Agent header value. #### Example 1 ```json {"input":{"user_agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36"}} ``` Result: ```json {"browser":{"name":"Chrome","version":"126.0.0.0"},"engine":{"name":"Blink","version":"126.0.0.0"},"os":{"name":"Windows","version":"10 or 11"},"device":"desktop","is_bot":false,"bot":null} ``` #### Example 2: An AI crawler ```json {"input":{"user_agent":"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)"}} ``` Result: ```json {"browser":{"name":null,"version":null},"engine":{"name":"WebKit","version":"537.36"},"os":{"name":null,"version":null},"device":"bot","is_bot":true,"bot":{"name":"ClaudeBot","kind":"ai","version":"1.0"}} ``` #### Example 3: Safari on an iPhone ```json {"input":{"user_agent":"Mozilla/5.0 (iPhone; CPU iPhone OS 17_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Mobile/15E148 Safari/604.1"}} ``` Result: ```json {"browser":{"name":"Safari","version":"17.5"},"engine":{"name":"WebKit","version":"605.1.15"},"os":{"name":"iOS","version":"17.5"},"device":"mobile","is_bot":false,"bot":null} ``` **Why does Windows say "10 or 11"?** Windows 11 still reports itself as "Windows NT 10.0", so the two cannot be told apart from this text. **Can a user agent be trusted?** No. Any client can send any text. Treat the result as a hint, not as proof of who is calling. **What are the bot kinds?** search, ai, social (link previews), seo, feed, monitor, tool (libraries and command-line clients) and other. ## Generators Make QR codes, charts, placeholders, calendar files and other ready-to-use output. ### .gitignore generator Build a .gitignore for the languages, editors and operating systems you use, merged without duplicates. - Endpoint: `POST https://setget.net/v1/tools/gitignore-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `gitignore-generate` - Page: https://setget.net/tools/gitignore-generate #### Inputs - `stacks` (required, array): What the project uses: python, node, java, go, rust, dotnet, php, ruby, cpp, swift, terraform, macos, windows, linux, vscode, jetbrains, vim, secrets. #### Example 1 ```json {"input":{"stacks":["rust","macos"]}} ``` Result: ```json {"text":"# Rust\ntarget/\n**/*.rs.bk\n*.pdb\n\n# macOS\n.DS_Store\n.AppleDouble\n.LSOverride\n._*\n.Spotlight-V100\n.Trashes\n.fseventsd\n","stacks":["rust","macos"],"patterns":10} ``` #### Example 2: A pattern both sections share (.env) is written once ```json {"input":{"stacks":["python","secrets"]}} ``` **A file is still tracked after I added it. Why?** .gitignore only affects files Git is not tracking yet. Run "git rm --cached " to stop tracking one that was already committed. **Where does the file go?** Save it as .gitignore in the top folder of the repository. ### ASCII banner text generator Turn short text into large five-line banner letters for terminals, READMEs, logs and code comments. - Endpoint: `POST https://setget.net/v1/tools/ascii-banner` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ascii-banner` - Page: https://setget.net/tools/ascii-banner #### Inputs - `text` (required, string): The text to enlarge: letters, digits and basic punctuation. Each line becomes its own banner row. - `character` (optional, string, default `"#"`): The character the letters are drawn with. - `spacing` (optional, integer, default `1`, range 0..4): Blank columns between letters. #### Example 1 ```json {"input":{"text":"HI"}} ``` Result: ```json {"text":"# # ###\n# # #\n##### #\n# # #\n# # ###","width":9,"height":5} ``` #### Example 2: Drawn with solid blocks ```json {"input":{"text":"v2.0","character":"█"}} ``` **Which characters are supported?** A to Z (lower case is drawn as capitals), 0 to 9 and common punctuation such as . , ! ? - _ : / + = ( ) # @ & *. **Why does my banner look crooked?** It needs a monospaced font. In Markdown, put it inside a code block. ### ASCII table generator Render headers and rows as a box-drawing or plain ASCII table for terminals, logs and code comments. - Endpoint: `POST https://setget.net/v1/tools/ascii-table` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ascii-table` - Page: https://setget.net/tools/ascii-table #### Inputs - `rows` (required, json): The rows: a list of lists, or a list of objects keyed by column name. - `headers` (optional, array): Column names. Leave out to use the object keys, or the first row. - `style` (optional, string, one of `box`, `rounded`, `double`, `ascii`, default `"box"`): Border style. ascii uses only + - and | for places that cannot show box-drawing characters. - `align_numbers` (optional, boolean, default `true`): Right-align columns that hold only numbers. - `max_width` (optional, integer, default `60`, range 5..200): Longest cell in characters; longer text is cut with an ellipsis. #### Example 1 ```json {"input":{"headers":["Name","Age"],"rows":[["Sam",31],["Alex",27]]}} ``` Result: ```json {"text":"┌──────┬─────┐\n│ Name │ Age │\n├──────┼─────┤\n│ Sam │ 31 │\n│ Alex │ 27 │\n└──────┴─────┘","rows":2,"columns":2,"width":14} ``` #### Example 2: Plain ASCII borders ```json {"input":{"headers":["Name","Age"],"rows":[["Sam",31],["Alex",27]],"style":"ascii"}} ``` Result: ```json {"text":"+------+-----+\n| Name | Age |\n+------+-----+\n| Sam | 31 |\n| Alex | 27 |\n+------+-----+","rows":2,"columns":2,"width":14} ``` **The borders look broken. Why?** Box-drawing characters need a monospaced font and UTF-8. Choose the ascii style where that is not available. ### Background pattern generator Make seamless SVG background patterns (dots, grid, stripes, checker or waves) with your colours and scale. - Endpoint: `POST https://setget.net/v1/tools/pattern-image` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `pattern-image` - Page: https://setget.net/tools/pattern-image #### Inputs - `pattern` (optional, string, one of `dots`, `grid`, `stripes`, `checker`, `waves`, default `"dots"`): The pattern to draw. - `foreground` (optional, string, default `"#cbd5e1"`): Pattern colour: hex or a basic colour name. - `background` (optional, string, default `"#f8fafc"`): Background colour: hex, a basic colour name, or "transparent". - `scale` (optional, integer, default `24`, range 4..400): Size of one repeating tile in pixels. - `width` (optional, integer, default `480`, range 1..4000): Width of the preview image in pixels. - `height` (optional, integer, default `320`, range 1..4000): Height of the preview image in pixels. #### Example 1: Dots ```json {"input":{"pattern":"dots"}} ``` #### Example 2: A small checkerboard ```json {"input":{"pattern":"checker","scale":8,"width":16,"height":16,"foreground":"#000","background":"#fff"}} ``` Result: ```json {"svg":"","width":16,"height":16,"tile":""} ``` #### Example 3: Waves on a dark background ```json {"input":{"pattern":"waves","scale":40,"foreground":"#2a78d6","background":"#0f172a"}} ``` **How do I use it as a page background?** Save the "tile" SVG as a file and set it as a repeating CSS background-image. The tile repeats without visible seams. ### Barcode generator Make a Code 128, EAN-13 or UPC-A barcode as SVG, with the check digit worked out or verified for you. - Endpoint: `POST https://setget.net/v1/tools/barcode-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `barcode-generate` - Page: https://setget.net/tools/barcode-generate #### Inputs - `text` (required, string): What to encode. Code 128 takes letters, digits and symbols; EAN-13 takes 12 or 13 digits; UPC-A takes 11 or 12. - `format` (optional, string, one of `code128`, `ean13`, `upca`, default `"code128"`): Barcode type. - `module_width` (optional, integer, default `2`, range 1..10): Width of the thinnest bar in pixels. - `height` (optional, integer, default `80`, range 10..1000): Height of the bars in pixels. - `show_text` (optional, boolean, default `true`): Print the value in readable characters under the bars. - `foreground` (optional, string, default `"#000000"`): Bar colour: hex or a basic colour name. - `background` (optional, string, default `"#ffffff"`): Background colour: hex, a basic colour name, or "transparent". #### Example 1: Code 128 ```json {"input":{"text":"SETGET-2026"}} ``` #### Example 2: EAN-13 with the check digit added (1) ```json {"input":{"text":"400638133393","format":"ean13"}} ``` #### Example 3: UPC-A ```json {"input":{"text":"03600029145","format":"upca","show_text":false}} ``` **Do I type the check digit?** Either way works. Leave it off and it is calculated; include it and it is verified, with the correct digit shown if yours is wrong. **Which type should I use?** Code 128 for labels, tickets and internal stock codes. EAN-13 and UPC-A are for retail products and need a number issued by GS1. ### Calendar event (.ics) generator Create a valid iCalendar (.ics) file for one event, with optional reminder and repeat rule, ready to import. - Endpoint: `POST https://setget.net/v1/tools/ics-event` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `ics-event` - Page: https://setget.net/tools/ics-event #### Inputs - `title` (required, string): Name of the event. - `start` (required, string): Start: a date (2026-03-14) for a whole-day event, or a date and time (2026-03-14T09:30). - `end` (optional, string): End, in the same form as the start. For whole-day events this is the last day. Leave out to use the duration. - `duration_minutes` (optional, integer, range 1..525600): Length in minutes, used when there is no end. Defaults to 60. - `time_zone` (optional, string, one of `utc`, `floating`, default `"utc"`): utc writes exact times in UTC (an offset such as +01:00 is converted). floating keeps the clock time in whatever zone the viewer is in, and ignores any offset. - `location` (optional, string): Where it takes place. - `description` (optional, string): Notes for the event. Line breaks are kept. - `url` (optional, string): A link for the event, such as a meeting address. - `reminder_minutes` (optional, integer, range 0..40320): Show a reminder this many minutes before the start. - `repeat` (optional, string, one of `none`, `daily`, `weekly`, `monthly`, `yearly`, default `"none"`): How often the event repeats. - `repeat_count` (optional, integer, range 1..1000): Total number of occurrences when repeating. - `repeat_until` (optional, string): Last date (2026-12-31) on which the event may repeat. Use this or repeat_count, not both. - `uid` (optional, string): Unique ID of the event. Derived from the content when left out, so the same input gives the same file. - `dtstamp` (optional, string): Creation time to record, as UTC (2026-03-01T12:00:00Z). Defaults to the event start. #### Example 1 ```json {"input":{"title":"Project kick-off","start":"2026-03-14T09:30","duration_minutes":45,"location":"Room 2, Main St","reminder_minutes":15}} ``` Result: ```json {"text":"BEGIN:VCALENDAR\r\nVERSION:2.0\r\nPRODID:-//SetGet//Event generator//EN\r\nCALSCALE:GREGORIAN\r\nMETHOD:PUBLISH\r\nBEGIN:VEVENT\r\nUID:87bf7650becd7028d0463f3a@setget.net\r\nDTSTAMP:20260314T093000Z\r\nDTSTART:20260314T093000Z\r\nDTEND:20260314T101500Z\r\nSUMMARY:Project kick-off\r\nLOCATION:Room 2\\, Main St\r\nBEGIN:VALARM\r\nACTION:DISPLAY\r\nDESCRIPTION:Project kick-off\r\nTRIGGER:-PT15M\r\nEND:VALARM\r\nEND:VEVENT\r\nEND:VCALENDAR\r\n","filename":"project-kick-off.ics","uid":"87bf7650becd7028d0463f3a@setget.net"} ``` #### Example 2: A whole-day event over five days, repeating yearly ```json {"input":{"title":"Team holiday","start":"2026-07-06","end":"2026-07-10","repeat":"yearly","repeat_count":3,"uid":"holiday-2026@example.com"}} ``` Result: ```json {"text":"BEGIN:VCALENDAR\r\nVERSION:2.0\r\nPRODID:-//SetGet//Event generator//EN\r\nCALSCALE:GREGORIAN\r\nMETHOD:PUBLISH\r\nBEGIN:VEVENT\r\nUID:holiday-2026@example.com\r\nDTSTAMP:20260706T000000Z\r\nDTSTART;VALUE=DATE:20260706\r\nDTEND;VALUE=DATE:20260711\r\nRRULE:FREQ=YEARLY;COUNT=3\r\nSUMMARY:Team holiday\r\nEND:VEVENT\r\nEND:VCALENDAR\r\n","filename":"team-holiday.ics","uid":"holiday-2026@example.com"} ``` **Which calendars open the file?** Google Calendar, Apple Calendar, Outlook and most others import .ics files. Save the text as event.ics. **Why is the end date one day later in the file?** iCalendar stores the end of a whole-day event as the day after the last day. You enter the last day; the file follows the standard. **What time zone are times in?** With "utc", a time without an offset is taken as UTC, and a time such as 09:30+01:00 is converted. With "floating", 09:30 means 09:30 wherever the viewer is. ### Chart generator Make a clean SVG bar, line, area, pie or donut chart from labels and values, with axes, legend and title. - Endpoint: `POST https://setget.net/v1/tools/chart` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `chart` - Page: https://setget.net/tools/chart #### Inputs - `type` (optional, string, one of `bar`, `hbar`, `line`, `area`, `pie`, `donut`, default `"bar"`): Chart type. hbar is a horizontal bar chart. - `labels` (optional, array): One label per value: the categories or points along the axis. Numbered 1, 2, 3 when left out. - `values` (optional, array): The numbers for a single series. - `series` (optional, json): Several series for bar, hbar, line or area charts: a list of {"name": ..., "values": [...]}, up to 8. - `title` (optional, string): Heading shown above the chart. - `width` (optional, integer, default `720`, range 240..2000): Width in pixels. - `height` (optional, integer, default `420`, range 160..2000): Height in pixels. - `theme` (optional, string, one of `light`, `dark`, default `"light"`): Light or dark colours. - `colours` (optional, array): Your own series colours in order, as hex or basic names. The built-in palette fills any gaps. - `show_values` (optional, boolean, default `false`): Print each value at the end of its bar (single-series bar charts, when there is room). #### Example 1: A bar chart ```json {"input":{"labels":["Mon","Tue","Wed","Thu","Fri"],"values":[12,19,7,15,22],"title":"Orders this week"}} ``` #### Example 2: Two lines on a dark background ```json {"input":{"type":"line","labels":["Q1","Q2","Q3","Q4"],"title":"Revenue by quarter","theme":"dark","series":[{"name":"2024","values":[31,42,38,51]},{"name":"2025","values":[36,47,52,64]}]}} ``` #### Example 3: A donut chart ```json {"input":{"type":"donut","labels":["Direct","Search","Social","Email"],"values":[420,310,150,90],"title":"Traffic sources"}} ``` **Why did my pie chart get an "Other" slice?** Past eight slices the colours stop being tellable apart, so the smallest slices are added together. Use a bar chart to show every item. **What happens with many labels?** Axis labels are tilted, shortened or thinned out so they never overlap. Hover over a bar or point in a browser to see its exact value. **Can I turn the SVG into a PNG?** Yes, with any SVG to PNG converter. SVG is best for web pages and documents because it stays sharp. ### Contact (vCard) QR code generator Make a QR code that adds a contact to a phone when scanned: name, phone, email, organisation and website. - Endpoint: `POST https://setget.net/v1/tools/qr-vcard` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `qr-vcard` - Page: https://setget.net/tools/qr-vcard #### Inputs - `name` (required, string): Full name of the contact. - `phone` (optional, string): Phone number, ideally in international form. - `email` (optional, string): Email address. - `organisation` (optional, string): Company or organisation. - `url` (optional, string): Website address, including https://. - `error_correction` (optional, string, one of `L`, `M`, `Q`, `H`, default `"M"`): How much damage the code survives: L about 7%, M 15%, Q 25%, H 30%. Higher levels hold less. - `module_size` (optional, integer, default `8`, range 1..40): Width of one small square (module) in pixels. - `margin` (optional, integer, default `4`, range 0..20): Blank quiet zone around the code, in modules. Scanners expect at least 4. - `foreground` (optional, string, default `"#000000"`): Colour of the dark squares: hex such as #000000, or a basic colour name. - `background` (optional, string, default `"#ffffff"`): Colour behind the code: hex, a basic colour name, or "transparent". #### Example ```json {"input":{"name":"Sam Rivera","phone":"+44 20 7946 0123","email":"sam@example.com","organisation":"Example Ltd","url":"https://example.com"}} ``` **How is the name split?** The last word is used as the family name and the rest as given names. Use the vCard tool with qr-generate if you need full control. **Why will my code not scan?** Keep strong contrast (dark squares on a light background), leave the margin at 4 or more, and print it large enough: about 2 cm across for a short link. Long text makes dense codes that need more size. ### Cron expression generator Build a cron expression from plain choices such as every 15 minutes or weekdays at 08:30, with an explanation. - Endpoint: `POST https://setget.net/v1/tools/crontab-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `crontab-generate` - Page: https://setget.net/tools/crontab-generate #### Inputs - `schedule` (required, string, one of `every_n_minutes`, `every_n_hours`, `daily`, `weekdays`, `weekly`, `monthly`): The kind of schedule. - `every` (optional, integer, default `15`, range 1..59): The N for every_n_minutes (1 to 59) or every_n_hours (1 to 23). - `time` (optional, string, default `"09:00"`): Time of day as HH:MM on a 24-hour clock, for daily, weekdays, weekly and monthly. - `day_of_week` (optional, string, one of `sunday`, `monday`, `tuesday`, `wednesday`, `thursday`, `friday`, `saturday`, default `"monday"`): The day for a weekly schedule. - `day_of_month` (optional, integer, default `1`, range 1..31): The day for a monthly schedule. #### Example 1 ```json {"input":{"schedule":"weekdays","time":"08:30"}} ``` Result: ```json {"expression":"30 8 * * 1-5","explanation":"At 08:30, Monday to Friday."} ``` #### Example 2 ```json {"input":{"schedule":"every_n_minutes","every":15}} ``` Result: ```json {"expression":"*/15 * * * *","explanation":"Every 15 minutes."} ``` #### Example 3: A day that not every month has ```json {"input":{"schedule":"monthly","day_of_month":31,"time":"23:00"}} ``` Result: ```json {"expression":"0 23 31 * *","explanation":"At 23:00 on day 31 of the month. Months with fewer days are skipped."} ``` **Which time zone does cron use?** The time zone of the machine that runs the job, often UTC on servers. Check before relying on an hour. **What do the five fields mean?** Minute, hour, day of month, month and day of week, in that order. A star means "every". **Why is "every 7 minutes" uneven?** Cron counts from minute 0 of each hour, so */7 runs at 0, 7 and so on to 56, then again at 0. Only steps that divide 60 are perfectly even. ### Emoji favicon generator Make an SVG favicon from an emoji or one or two letters, with the ready-to-paste link tag for your page head. - Endpoint: `POST https://setget.net/v1/tools/emoji-favicon` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `emoji-favicon` - Page: https://setget.net/tools/emoji-favicon #### Inputs - `text` (required, string): One emoji, or one or two letters or digits. - `background` (optional, string, default `"#1e293b"`): Colour of the square behind it: hex, a basic colour name, or "transparent". - `text_colour` (optional, string): Letter colour. Chosen for contrast when left out. Emoji keep their own colours. - `shape` (optional, string, one of `square`, `rounded`, `circle`, default `"rounded"`): Outline of the icon. - `size` (optional, integer, default `64`, range 16..512): Width and height of the SVG in pixels. Browsers scale it as needed. #### Example 1 ```json {"input":{"text":"S","background":"#2a78d6","shape":"square"}} ``` Result: ```json {"svg":"S","width":64,"height":64,"html":""} ``` #### Example 2: An emoji on a dark rounded square ```json {"input":{"text":"🚀"}} ``` **Where do I paste the tag?** Inside the of your page. No image file is needed: the icon lives in the tag itself. **Does every browser show it?** Current Chrome, Edge, Firefox and Opera show SVG favicons. Safari support is partial, so keep a favicon.ico as a fallback if that matters. ### Gradient image generator Make an SVG image filled with a linear or radial gradient from your colour stops, plus the matching CSS. - Endpoint: `POST https://setget.net/v1/tools/gradient-image` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `gradient-image` - Page: https://setget.net/tools/gradient-image #### Inputs - `colours` (required, array): Two to twelve colours in order. Add a position to place a stop, such as "#ff0080 30%". - `type` (optional, string, one of `linear`, `radial`, default `"linear"`): Gradient shape. - `angle` (optional, number, default `90`, range -360..360): Direction of a linear gradient in degrees, as in CSS: 0 runs bottom to top, 90 left to right. - `width` (optional, integer, default `800`, range 1..4000): Width in pixels. - `height` (optional, integer, default `400`, range 1..4000): Height in pixels. #### Example 1 ```json {"input":{"colours":["#2a78d6","#e87ba4"],"width":200,"height":100}} ``` Result: ```json {"svg":"","width":200,"height":100,"css":"linear-gradient(90deg, #2a78d6 0%, #e87ba4 100%)"} ``` #### Example 2: A radial gradient with a placed stop ```json {"input":{"colours":["#fde68a","#f97316 40%","#7c2d12"],"type":"radial"}} ``` **How does the angle work?** As in CSS: 0 points up, 90 points right, 180 points down. The "css" value in the result gives the same gradient for a stylesheet. ### Identicon and avatar generator Turn any name, email or ID into a unique avatar: a symmetric 5×5 identicon or coloured initials, as SVG. - Endpoint: `POST https://setget.net/v1/tools/identicon` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `identicon` - Page: https://setget.net/tools/identicon #### Inputs - `text` (required, string): The name, email address or ID to draw. The same text always gives the same image. - `style` (optional, string, one of `blocks`, `initials`, default `"blocks"`): blocks draws a symmetric pattern; initials draws one or two letters. - `size` (optional, integer, default `240`, range 16..2000): Width and height in pixels. - `shape` (optional, string, one of `square`, `rounded`, `circle`, default `"rounded"`): Outline of the image. - `background` (optional, string): Background colour. Defaults to light grey for blocks and the generated colour for initials. #### Example 1: A block identicon ```json {"input":{"text":"sam.rivera@example.com"}} ``` #### Example 2: Initials in a circle ```json {"input":{"text":"Sam Rivera","style":"initials","shape":"circle","size":96}} ``` **Is it the same as a GitHub identicon?** It is the same idea (a hash drawn as a mirrored 5×5 grid) but not the same picture. It is stable here: one text always gives one image. **How are initials chosen?** The first letters of the first two words. For an email address only the part before @ is used. ### Markdown table generator Build a Markdown table from headers and rows, padded so the columns also line up as plain text. - Endpoint: `POST https://setget.net/v1/tools/markdown-table` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `markdown-table` - Page: https://setget.net/tools/markdown-table #### Inputs - `rows` (required, json): The rows: a list of lists, or a list of objects keyed by column name. - `headers` (optional, array): Column names. Leave out to use the object keys, or the first row. - `align` (optional, array): Alignment: left, center, right or none. One value for every column, or one per column. #### Example 1 ```json {"input":{"headers":["Name","Age"],"rows":[["Sam",31],["Alex",27]],"align":["left","right"]}} ``` Result: ```json {"text":"| Name | Age |\n| :--- | --: |\n| Sam | 31 |\n| Alex | 27 |","rows":2,"columns":2} ``` #### Example 2: Objects as rows; the pipe character is escaped ```json {"input":{"rows":[{"tool":"a|b","ok":true}]}} ``` Result: ```json {"text":"| tool | ok |\n| ---- | ---- |\n| a\\|b | true |","rows":1,"columns":2} ``` **Why pad the cells?** Markdown ignores the extra spaces, but padded tables are far easier to read and edit as plain text. ### Open source licence generator Get the text of the MIT, BSD, ISC, Apache-2.0 or Unlicense licence with your year and name filled in. - Endpoint: `POST https://setget.net/v1/tools/license-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `license-generate` - Page: https://setget.net/tools/license-generate #### Inputs - `license` (optional, string, one of `MIT`, `ISC`, `BSD-2-Clause`, `BSD-3-Clause`, `Apache-2.0`, `Unlicense`, default `"MIT"`): The licence, by its SPDX name. Apache-2.0 returns the short notice, not the full licence text. - `holder` (optional, string): Copyright holder: your name or organisation. Not needed for the Unlicense. - `year` (optional, integer, range 1900..2200): Copyright year, usually the year of first publication. Not needed for the Unlicense. #### Example 1 ```json {"input":{"license":"ISC","holder":"Sam Rivera","year":2026}} ``` Result: ```json {"text":"ISC License\n\nCopyright (c) 2026 Sam Rivera\n\nPermission to use, copy, modify, and/or distribute this software for any\npurpose with or without fee is hereby granted, provided that the above\ncopyright notice and this permission notice appear in all copies.\n\nTHE SOFTWARE IS PROVIDED \"AS IS\" AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH\nREGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY\nAND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT,\nINDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM\nLOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR\nOTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR\nPERFORMANCE OF THIS SOFTWARE.\n","license":"ISC","filename":"LICENSE","complete":true} ``` #### Example 2: MIT ```json {"input":{"license":"MIT","holder":"Example Ltd","year":2026}} ``` #### Example 3: Public domain dedication; no name needed ```json {"input":{"license":"Unlicense"}} ``` **Which licence should I pick?** MIT and ISC are short and permissive. BSD-3-Clause adds a rule against using your name for endorsement. Apache-2.0 adds a patent grant. The Unlicense gives the work to the public domain. This is not legal advice. **Why is the year not filled in for me?** The tool gives the same answer for the same input, so it never reads the clock. Enter the year. **Why is Apache-2.0 not the full text?** The full licence is about 11,000 characters and must be copied exactly, so the tool returns the official short notice and tells you where to download the full text. ### Placeholder image generator Make an SVG placeholder image of any size with a centred label, for mock-ups and layouts. - Endpoint: `POST https://setget.net/v1/tools/placeholder-image` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `placeholder-image` - Page: https://setget.net/tools/placeholder-image #### Inputs - `width` (optional, integer, default `600`, range 1..4000): Width in pixels. - `height` (optional, integer, default `400`, range 1..4000): Height in pixels. - `label` (optional, string): Text to show. Defaults to the size, such as 600×400. - `background` (optional, string, default `"#e2e8f0"`): Background colour: hex or a basic colour name. - `text_colour` (optional, string, default `"#64748b"`): Label colour: hex or a basic colour name. #### Example 1 ```json {"input":{"width":120,"height":60}} ``` Result: ```json {"svg":"120×60","width":120,"height":60} ``` #### Example 2: A custom label and colours ```json {"input":{"width":800,"height":200,"label":"Hero banner","background":"#1e293b","text_colour":"#f8fafc"}} ``` **Can I use it as an img source?** Yes. Save the SVG as a file, or call the API URL and use the "svg" value. SVG stays sharp at any size. ### Progress bar image generator Make an SVG progress bar with a percentage label, your colours, rounded corners and optional stripes. - Endpoint: `POST https://setget.net/v1/tools/progress-bar-image` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `progress-bar-image` - Page: https://setget.net/tools/progress-bar-image #### Inputs - `value` (required, number): Current progress. - `maximum` (optional, number, default `100`): The value that means complete. - `width` (optional, integer, default `320`, range 20..4000): Width in pixels. - `height` (optional, integer, default `28`, range 4..400): Height in pixels. - `colour` (optional, string, default `"#1baf7a"`): Colour of the filled part: hex or a basic colour name. - `background` (optional, string, default `"#e5e7eb"`): Colour of the empty track. - `label` (optional, string): Text to show instead of the percentage. - `show_label` (optional, boolean, default `true`): Show the percentage (or your label) on the bar. - `radius` (optional, integer, range 0..200): Corner radius in pixels. Leave out for fully rounded ends; 0 gives square corners. - `striped` (optional, boolean, default `false`): Overlay diagonal stripes on the filled part. #### Example 1: 72% ```json {"input":{"value":72}} ``` #### Example 2: A custom label with stripes ```json {"input":{"value":3,"maximum":8,"label":"3 of 8 done","colour":"#2a78d6","striped":true,"radius":4}} ``` **What if the value is over the maximum?** The bar is drawn full, and the percentage label is capped at 100%. Values below zero show as 0%. ### QR code generator Make a scannable QR code as SVG from any text or link, with a choice of size, colours and error correction. A complete QR encoder (byte mode, UTF-8, versions 1 to 40) that picks the smallest version that fits, tries all eight masks and reads every code back before returning it. - Endpoint: `POST https://setget.net/v1/tools/qr-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `qr-generate` - Page: https://setget.net/tools/qr-generate #### Inputs - `text` (required, string): The text or link to store in the code. - `error_correction` (optional, string, one of `L`, `M`, `Q`, `H`, default `"M"`): How much damage the code survives: L about 7%, M 15%, Q 25%, H 30%. Higher levels hold less. - `module_size` (optional, integer, default `8`, range 1..40): Width of one small square (module) in pixels. - `margin` (optional, integer, default `4`, range 0..20): Blank quiet zone around the code, in modules. Scanners expect at least 4. - `foreground` (optional, string, default `"#000000"`): Colour of the dark squares: hex such as #000000, or a basic colour name. - `background` (optional, string, default `"#ffffff"`): Colour behind the code: hex, a basic colour name, or "transparent". #### Example 1: A link ```json {"input":{"text":"https://setget.net"}} ``` #### Example 2: High error correction in dark blue ```json {"input":{"text":"Table 12","error_correction":"H","foreground":"#1e3a8a","module_size":6}} ``` **How much text fits?** Up to 2,953 bytes at level L and 1,273 at level H. Short content gives a simpler code that scans faster, so prefer a short link. **Does the code expire or track scans?** No. The text is stored in the picture itself. Nothing is redirected or counted. **Why will my code not scan?** Keep strong contrast (dark squares on a light background), leave the margin at 4 or more, and print it large enough: about 2 cm across for a short link. Long text makes dense codes that need more size. ### Redirect rules generator Turn a list of old and new addresses into redirect rules for Apache .htaccess, Nginx, or a Netlify _redirects file. - Endpoint: `POST https://setget.net/v1/tools/htaccess-redirects` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `htaccess-redirects` - Page: https://setget.net/tools/htaccess-redirects #### Inputs - `redirects` (required, json): The redirects: a list of {"from": "/old", "to": "/new"} objects or ["/old", "/new"] pairs. Plain text with "old new" on each line also works. - `format` (optional, string, one of `apache`, `nginx`, `netlify`, default `"apache"`): Which server the rules are for. netlify is the _redirects format, also used by Cloudflare Pages. - `status` (optional, integer, one of `301`, `302`, default `301`): 301 for a permanent move (best for SEO), 302 for a temporary one. #### Example 1 ```json {"input":{"redirects":[{"from":"/old-page","to":"/new-page"},["/blog/2019/post.html","https://blog.example.com/post"]]}} ``` Result: ```json {"text":"RedirectMatch 301 ^/old-page$ /new-page\nRedirectMatch 301 ^/blog/2019/post\\.html$ https://blog.example.com/post\n","count":2,"filename":".htaccess"} ``` #### Example 2: Nginx, from plain text ```json {"input":{"redirects":"/old-page /new-page\n/sale /offers","format":"nginx","status":302}} ``` Result: ```json {"text":"location = /old-page { return 302 /new-page; }\nlocation = /sale { return 302 /offers; }\n","count":2,"filename":"redirects.conf"} ``` #### Example 3: Netlify and Cloudflare Pages ```json {"input":{"redirects":[["/old-page","/new-page"]],"format":"netlify"}} ``` Result: ```json {"text":"/old-page /new-page 301\n","count":1,"filename":"_redirects"} ``` **Where do the rules go?** Apache: in the .htaccess file at the site root. Nginx: inside the server { } block. Netlify and Cloudflare Pages: in a file named _redirects in the published folder. **Do the rules match the exact address?** Yes. Each rule matches that one path, not everything beneath it. Query strings (the part after ?) are not part of the match. **301 or 302?** Use 301 when the move is permanent so search engines transfer ranking to the new address. Use 302 for short-term moves. ### robots.txt generator Build a robots.txt: allow or block all crawlers, add your own rules, block AI crawlers and list sitemaps. - Endpoint: `POST https://setget.net/v1/tools/robots-txt-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `robots-txt-generate` - Page: https://setget.net/tools/robots-txt-generate #### Inputs - `mode` (optional, string, one of `allow_all`, `block_all`, `custom`, default `"allow_all"`): allow_all lets crawlers visit everything, block_all shuts them out, custom uses your rules. - `rules` (optional, json): For custom mode: a list of {"user_agent": "*", "disallow": ["/admin/"], "allow": ["/admin/help"], "crawl_delay": 5}. - `block_ai` (optional, boolean, default `false`): Add a rule that blocks common AI crawlers: GPTBot, ClaudeBot, CCBot, Google-Extended, PerplexityBot. - `sitemaps` (optional, array): Full addresses of your sitemap files. - `crawl_delay` (optional, number, range 0..3600): Seconds crawlers should wait between requests, for the general rule. Google ignores this. #### Example 1 ```json {"input":{"mode":"allow_all","block_ai":true,"sitemaps":["https://example.com/sitemap.xml"]}} ``` Result: ```json {"text":"User-agent: *\nDisallow:\n\nUser-agent: GPTBot\nUser-agent: ClaudeBot\nUser-agent: CCBot\nUser-agent: Google-Extended\nUser-agent: PerplexityBot\nDisallow: /\n\nSitemap: https://example.com/sitemap.xml\n","groups":2} ``` #### Example 2: Custom rules for two crawlers ```json {"input":{"mode":"custom","rules":[{"user_agent":"*","disallow":["/admin/","/cart"],"allow":["/admin/help"]},{"user_agent":"Bingbot","disallow":["/"],"crawl_delay":10}]}} ``` Result: ```json {"text":"User-agent: *\nAllow: /admin/help\nDisallow: /admin/\nDisallow: /cart\n\nUser-agent: Bingbot\nDisallow: /\nCrawl-delay: 10\n","groups":2} ``` **Does robots.txt hide pages from search results?** No. It asks crawlers not to fetch pages. A blocked page can still be listed if other sites link to it. Use a noindex tag or a login to keep a page out of results. **Do AI crawlers obey it?** The named crawlers say they do, but robots.txt is a request, not a lock. **Where does the file go?** At the root of the site, for example https://example.com/robots.txt. ### Sample data generator Generate fake but realistic sample records (names, emails, phones, addresses and more) as JSON for testing. - Endpoint: `POST https://setget.net/v1/tools/lorem-data` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `lorem-data` - Page: https://setget.net/tools/lorem-data #### Inputs - `fields` (optional, array, default `["name", "email"]`): Which fields each record has, in order: id, name, email, phone, address, company, date, sentence. - `count` (optional, integer, default `10`, range 1..500): How many records to make. - `seed` (optional, integer, default `1`, range -1000000000000..1000000000000): Any whole number. The same seed always gives the same records; change it for a different set. #### Example 1: Two people ```json {"input":{"fields":["name","email","phone"],"count":2,"seed":7}} ``` Result: ```json {"text":"[\n {\n \"name\": \"Liam Evans\",\n \"email\": \"liam.evans@example.com\",\n \"phone\": \"+1 512 555 0183\"\n },\n {\n \"name\": \"Ben Chen\",\n \"email\": \"ben.chen@example.com\",\n \"phone\": \"+1 702 555 0112\"\n }\n]","records":[{"name":"Liam Evans","email":"liam.evans@example.com","phone":"+1 512 555 0183"},{"name":"Ben Chen","email":"ben.chen@example.com","phone":"+1 702 555 0112"}],"count":2} ``` #### Example 2: IDs, companies and dates ```json {"input":{"fields":["id","company","date"],"count":3,"seed":42}} ``` **Could the data belong to a real person?** It is built to avoid that. Emails use example.com, which is reserved for examples, phone numbers use the 555-01xx range kept for fiction, and addresses use invented towns. **Why do I get the same records every time?** The seed makes the output repeatable, which is what tests need. Change the seed for new records. ### Social preview card generator Make a 1200×630 social preview (Open Graph) card as SVG with a wrapped title, subtitle and site name. - Endpoint: `POST https://setget.net/v1/tools/og-card` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `og-card` - Page: https://setget.net/tools/og-card #### Inputs - `title` (required, string): The headline. It wraps to at most three lines and shrinks to fit. - `subtitle` (optional, string): A smaller line under the title. - `site_name` (optional, string): Your site or author name, shown at the bottom. - `accent` (optional, string, default `"#2a78d6"`): Accent colour: hex or a basic colour name. - `theme` (optional, string, one of `dark`, `light`, default `"dark"`): Dark or light background. #### Example 1 ```json {"input":{"title":"How we cut our build time from 14 minutes to 90 seconds","subtitle":"Notes from a year of small fixes","site_name":"example.com"}} ``` #### Example 2: A short title on a light card ```json {"input":{"title":"Release 2.0","theme":"light","accent":"#eb6834"}} ``` **Can I use the SVG directly as og:image?** Most social networks need PNG or JPEG. Convert the SVG to PNG (1200×630) before using it in an og:image tag. ### Sparkline generator Draw a tiny line or bar chart as SVG from a list of numbers, small enough to sit in a table cell or README. - Endpoint: `POST https://setget.net/v1/tools/sparkline` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `sparkline` - Page: https://setget.net/tools/sparkline #### Inputs - `values` (required, array): The numbers to plot, in order. - `type` (optional, string, one of `line`, `bar`, default `"line"`): Chart style. - `width` (optional, integer, default `160`, range 20..2000): Width in pixels. - `height` (optional, integer, default `40`, range 10..1000): Height in pixels. - `colour` (optional, string, default `"#2a78d6"`): Line or bar colour: hex or a basic colour name. - `fill` (optional, boolean, default `false`): Shade the area under a line. - `markers` (optional, boolean, default `false`): Mark the lowest and highest values. - `stroke_width` (optional, number, default `2`, range 0.5..10): Line thickness in pixels. #### Example 1 ```json {"input":{"values":[1,3,2]}} ``` Result: ```json {"svg":"","width":160,"height":40,"min":1,"max":3,"last":2} ``` #### Example 2: Shaded, with markers ```json {"input":{"values":[3,7,4,9,6,12,8],"fill":true,"markers":true}} ``` #### Example 3: Bars with negative values ```json {"input":{"values":[5,-2,4,8,-6,3],"type":"bar","colour":"#1baf7a"}} ``` **What is a sparkline?** A word-sized chart with no axes or labels. It shows the shape of a trend, not exact values. ### Table image generator Render a small table as a tidy SVG image with a header row, zebra stripes and right-aligned numbers. - Endpoint: `POST https://setget.net/v1/tools/table-image` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `table-image` - Page: https://setget.net/tools/table-image #### Inputs - `rows` (required, json): The rows: a list of lists, or a list of objects keyed by column name. Up to 30 rows and 10 columns. - `headers` (optional, array): Column names. Leave out to use the object keys, or the first row. - `font_size` (optional, integer, default `14`, range 8..32): Text size in pixels. - `theme` (optional, string, one of `light`, `dark`, default `"light"`): Light or dark colours. #### Example 1 ```json {"input":{"headers":["City","Temp"],"rows":[["Oslo",4],["Rome",18]]}} ``` #### Example 2: Rows as objects, dark theme ```json {"input":{"rows":[{"name":"Sam","role":"Admin"},{"name":"Alex","role":"Editor"}],"theme":"dark"}} ``` **What happens to long cells?** Each column is as wide as its longest cell up to a limit; longer text is cut with an ellipsis. ### vCard generator Create a vCard (.vcf) contact file, version 3.0 or 4.0, with fields properly escaped and long lines folded. - Endpoint: `POST https://setget.net/v1/tools/vcard` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `vcard` - Page: https://setget.net/tools/vcard #### Inputs - `first_name` (optional, string): Given name. - `last_name` (optional, string): Family name. - `organisation` (optional, string): Company or organisation. - `job_title` (optional, string): Job title or role. - `phone` (optional, string): Phone number, ideally in international form. - `phone_type` (optional, string, one of `cell`, `work`, `home`, default `"cell"`): Kind of phone number. - `email` (optional, string): Email address. - `url` (optional, string): Website address, including https://. - `street` (optional, string): Street address. - `city` (optional, string): Town or city. - `region` (optional, string): County, state or province. - `postcode` (optional, string): Postcode or ZIP code. - `country` (optional, string): Country. - `note` (optional, string): Free-text note. Line breaks are kept. - `version` (optional, string, one of `3.0`, `4.0`, default `"3.0"`): vCard version. 3.0 is the most widely supported. #### Example 1 ```json {"input":{"first_name":"Sam","last_name":"Rivera","organisation":"Example Ltd","phone":"+44 20 7946 0123","email":"sam@example.com"}} ``` Result: ```json {"text":"BEGIN:VCARD\r\nVERSION:3.0\r\nN:Rivera;Sam;;;\r\nFN:Sam Rivera\r\nORG:Example Ltd\r\nTEL;TYPE=CELL:+44 20 7946 0123\r\nEMAIL;TYPE=INTERNET:sam@example.com\r\nEND:VCARD\r\n","filename":"sam-rivera.vcf"} ``` #### Example 2: Version 4.0 with escaped punctuation ```json {"input":{"first_name":"Alex","organisation":"Smith, Jones; Partners","city":"Leeds","country":"United Kingdom","version":"4.0"}} ``` Result: ```json {"text":"BEGIN:VCARD\r\nVERSION:4.0\r\nN:;Alex;;;\r\nFN:Alex\r\nORG:Smith\\, Jones\\; Partners\r\nADR:;;;Leeds;;;United Kingdom\r\nEND:VCARD\r\n","filename":"alex.vcf"} ``` **How do I use the result?** Save the text as a file ending in .vcf and open it on a phone or in a mail program to add the contact. **Which version should I choose?** Use 3.0 unless you know the receiving system wants 4.0. Every common address book reads 3.0. ### WiFi QR code generator Make a QR code that joins a WiFi network when scanned, with the network name and password escaped correctly. - Endpoint: `POST https://setget.net/v1/tools/qr-wifi` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `qr-wifi` - Page: https://setget.net/tools/qr-wifi #### Inputs - `ssid` (required, string): The network name, exactly as it appears (capitals matter). - `password` (optional, string): The network password. Leave empty for an open network. - `security` (optional, string, one of `WPA`, `WEP`, `nopass`, default `"WPA"`): Network security. WPA covers WPA2 and WPA3. Use nopass for an open network. - `hidden` (optional, boolean, default `false`): The network does not broadcast its name. - `error_correction` (optional, string, one of `L`, `M`, `Q`, `H`, default `"M"`): How much damage the code survives: L about 7%, M 15%, Q 25%, H 30%. Higher levels hold less. - `module_size` (optional, integer, default `8`, range 1..40): Width of one small square (module) in pixels. - `margin` (optional, integer, default `4`, range 0..20): Blank quiet zone around the code, in modules. Scanners expect at least 4. - `foreground` (optional, string, default `"#000000"`): Colour of the dark squares: hex such as #000000, or a basic colour name. - `background` (optional, string, default `"#ffffff"`): Colour behind the code: hex, a basic colour name, or "transparent". #### Example 1: A WPA network ```json {"input":{"ssid":"Cafe Guest","password":"latte;art"}} ``` #### Example 2: An open, hidden network ```json {"input":{"ssid":"Open Net","security":"nopass","hidden":true}} ``` **Is my password sent anywhere else?** It is only used to draw the code. Anyone who can see the code can read the password, so display it only where guests should have it. **Which phones can scan it?** The cameras of current iPhone and Android phones read this format and offer to join the network. ### XML sitemap generator Build a valid XML sitemap from a list of page addresses, with optional last-modified date, frequency and priority. - Endpoint: `POST https://setget.net/v1/tools/sitemap-generate` with JSON `{"input": {...}}` - MCP: `run_tool` with `slug` = `sitemap-generate` - Page: https://setget.net/tools/sitemap-generate #### Inputs - `urls` (required, json): The pages: a list of full addresses, or of objects {"loc": ..., "lastmod": ..., "changefreq": ..., "priority": ...}. Plain text with one address per line also works. - `lastmod` (optional, string): Last-modified date (2026-03-14) for pages that do not give their own. - `changefreq` (optional, string, one of `always`, `hourly`, `daily`, `weekly`, `monthly`, `yearly`, `never`): How often pages change, for pages that do not give their own. - `priority` (optional, number, range 0..1): Priority from 0.0 to 1.0, for pages that do not give their own. #### Example 1 ```json {"input":{"urls":["https://example.com/","https://example.com/search?q=a&page=2"],"lastmod":"2026-03-14"}} ``` Result: ```json {"text":"\n\n \n https://example.com/\n 2026-03-14\n \n \n https://example.com/search?q=a&page=2\n 2026-03-14\n \n\n","urls":2,"duplicates_removed":0} ``` #### Example 2: Per-page details ```json {"input":{"urls":[{"loc":"https://example.com/","priority":1.0,"changefreq":"daily"},{"loc":"https://example.com/blog/","lastmod":"2026-02-01"}]}} ``` **How many pages can a sitemap hold?** The standard allows 50,000 per file. This tool builds up to 5,000 at a time. **Do search engines use changefreq and priority?** Google ignores both and uses lastmod only when it is accurate. They are optional, so leave them out unless another system needs them. ## Files and archives Checksum files, and pack and unpack zip archives. ### Create a zip archive Pack up to 20 uploaded files into one zip archive, with a choice of compression level. - Endpoint: `POST https://setget.net/v1/tools/zip-create` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `zip-create` - Page: https://setget.net/tools/zip-create #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `files` (required, files, a list of file references such as `["DROP_ID/f1"]`, at most 20): The files to pack. - `name` (optional, string, default `"archive.zip"`): File name of the archive. - `level` (optional, integer, default `6`, range 0..9): Compression level: 0 stores without compressing, 9 is smallest and slowest. #### Example: Zip two files Upload first. Each placeholder below stands for the reference of one upload: `DROP_1/f1` = [sample.txt](https://setget.net/static/samples/sample.txt), `DROP_2/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"files":["DROP_1/f1","DROP_2/f1"]},"wait":30} ``` ### Extract a zip archive Unpack a zip archive into separate files. Guards against zip bombs; programs inside are skipped. - Endpoint: `POST https://setget.net/v1/tools/zip-extract` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `zip-extract` - Page: https://setget.net/tools/zip-extract #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `archive` (required, file, a file reference such as `DROP_ID/f1`, accepts zip): The zip archive to unpack. #### Example: Unpack an archive Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.zip](https://setget.net/static/samples/sample.zip). ```json {"input":{"archive":"DROP_ID/f1"},"wait":30} ``` **What are the limits?** Up to 100 files and 200 MB unpacked. Password-protected archives are not supported. **What happens to folders?** Files keep their names, with the folder path joined in by underscores, for example docs_readme.txt. ### File checksum Compute MD5, SHA-1, SHA-256, SHA-512 and CRC32 checksums of a file to verify a download or compare files. - Endpoint: `POST https://setget.net/v1/tools/file-checksum` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `file-checksum` - Page: https://setget.net/tools/file-checksum #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `file` (required, file, a file reference such as `DROP_ID/f1`): The file to checksum. - `algorithms` (optional, array, default `["sha256"]`): Which checksums to compute: md5, sha1, sha256, sha512, crc32. #### Example 1: SHA-256 of a text file Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.txt](https://setget.net/static/samples/sample.txt). ```json {"input":{"file":"DROP_ID/f1"},"wait":30} ``` #### Example 2: MD5 and CRC32 Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"file":"DROP_ID/f1","algorithms":["md5","crc32"]},"wait":30} ``` **Which checksum should I use?** SHA-256 for verifying downloads and detecting tampering. MD5 and CRC32 only catch accidental corruption; they are not safe against deliberate changes. ### File type and info Identify what a file really is from its contents: type, media type, size and whether its name matches. - Endpoint: `POST https://setget.net/v1/tools/file-info` with JSON `{"input": {...}}` (add `"wait": 30` to hold the reply until the job finishes) - MCP: `run_tool` with `slug` = `file-info` - Page: https://setget.net/tools/file-info #### Files in, files out This is a file tool. Its file inputs take file references, not file contents and not web addresses. - Get a reference by uploading: `curl -T FILE https://setget.net/drop/` prints a line `ref /f1`. Or `POST https://setget.net/v1/drops` with `{"files":[{"name","size"}]}`, PUT the bytes to the returned `upload.url` with the listed headers, and use the file's `ref`. Over MCP use the `drop` tool. A file under about 700 KB can also be sent in place as a `data:;base64,` URI. - Uploads are at most 100 MB per file, are readable by anyone with the link, and expire (one hour by default). Programs are refused. - The tool runs as a job. The reply is `{"tool", "ms", "result", "files", "job"}`: `result` is JSON, and every result file comes back in `files` as `{"name", "size", "content_type", "kind", "url", "ref"}`. Download it from `url`, or pass `ref` as the input of another file tool. - A job still running when the wait ends answers 202 with `job`. Follow it with `GET https://setget.net/v1/jobs/?wait=30`, or stream progress from `GET https://setget.net/v1/jobs//events` (over MCP: the `job` tool). A person can watch it at `job.url`. #### Inputs - `file` (required, file, a file reference such as `DROP_ID/f1`): The file to inspect. #### Example: Inspect an image Upload first. Each placeholder below stands for the reference of one upload: `DROP_ID/f1` = [sample.png](https://setget.net/static/samples/sample.png). ```json {"input":{"file":"DROP_ID/f1"},"wait":30} ``` **How is the type detected?** From the first bytes of the file, not from its name. A video renamed to .png is still reported as a video. ## Error codes ### unknown_tool (HTTP 404) No tool has that name. There is no tool with that name. What to do: Check the spelling, or search the catalogue at https://setget.net/tools or GET /v1/tools/search?q=... Safe to retry unchanged: no. ### invalid_input (HTTP 400) The input is not valid. One or more inputs are missing or not valid. What to do: Fix the fields listed under "problems" and send the request again. Each tool page lists its inputs with examples. Safe to retry unchanged: no. ### tool_timeout (HTTP 422) The tool took too long. This input took longer to process than the time limit allows. What to do: Try a smaller input, or a simpler pattern or expression. Safe to retry unchanged: no. ### tool_memory (HTTP 422) The tool ran out of memory. This input needs more memory than a single tool call is allowed to use. What to do: Try a smaller input, or split it into parts and process each part separately. Safe to retry unchanged: no. ### output_too_large (HTTP 422) The result is too large to return. The result is larger than the 1 MiB response limit. What to do: Ask for less output, for example fewer items or a smaller count, or split the input. Safe to retry unchanged: no. ### tool_busy (HTTP 503) All tool workers are busy. Every tool worker is busy right now. What to do: Wait a second and try again. Nothing was changed, so retrying is safe. Safe to retry unchanged: yes. ### tool_disabled (HTTP 503) This tool is switched off. This tool is temporarily switched off. What to do: Try again later, or look for a related tool at https://setget.net/tools. Safe to retry unchanged: yes. ### tool_failed (HTTP 500) The tool hit an unexpected problem. Something went wrong on our side while running this tool. What to do: Try again. If it keeps happening, send the request_id to https://setget.net/feedback so we can find the cause. Safe to retry unchanged: yes. ### tool_daily_limit (HTTP 429) Daily tool limit reached. You have used all of today's free tool calls from this network address. What to do: Wait for the limit to reset. The message says when that is. Safe to retry unchanged: no. ### storage_off (HTTP 503) File tools are not switched on. File uploads and file tools are not available on this server yet. What to do: Use the tools that work on text input for now, or try again later. Safe to retry unchanged: no. ### storage_unavailable (HTTP 503) File storage is temporarily unavailable. The service could not reach its file storage. What to do: Try again in a few seconds. Uploads that already finished are not affected. Safe to retry unchanged: yes. ### file_too_large (HTTP 413) The file is too large. This file is larger than the 100 MB limit. What to do: Make the file smaller first, for example by compressing or trimming it, then upload it again. Safe to retry unchanged: no. ### too_many_files (HTTP 400) Too many files. A drop can hold at most 20 files. What to do: Split the files across more than one drop, or zip them into one file first. Safe to retry unchanged: no. ### file_not_found (HTTP 404) That file reference does not point at a file. No file was found for that reference. What to do: Use the "ref" value returned when the file was uploaded, for example "DROP_ID/f1". Files expire, so upload it again if it is gone. Safe to retry unchanged: no. ### upload_incomplete (HTTP 409) The upload has not arrived yet. This file was announced but its bytes have not been uploaded yet. What to do: Send the file to its upload URL with PUT first, then try again. Safe to retry unchanged: yes. ### file_refused (HTTP 422) This kind of file is not accepted. Programs and installers cannot be uploaded. What to do: Upload documents, images, audio, video, archives or text instead. Safe to retry unchanged: no. ### unsupported_file (HTTP 422) This tool cannot read that file. The file is not a type this tool can read. What to do: Check the tool page for the file types it accepts, and convert the file first if needed. Safe to retry unchanged: no. ### upload_daily_limit (HTTP 429) Daily upload limit reached. You have reached today's free upload allowance from this network address. What to do: Wait for the limit to reset. The message says when that is. Safe to retry unchanged: no. ### job_not_found (HTTP 404) No job has that ID. There is no job with that ID. It may have expired. What to do: Check the ID. Jobs and their results expire, so run the tool again if it is gone. Safe to retry unchanged: no. ### job_daily_limit (HTTP 429) Daily job limit reached. You have used all of today's free file jobs from this network address. What to do: Wait for the limit to reset. The message says when that is. Safe to retry unchanged: no. ### too_many_jobs (HTTP 429) Too many jobs waiting. You already have the maximum number of jobs waiting or running. What to do: Wait for one of your jobs to finish, or cancel one, then try again. Safe to retry unchanged: yes. ### jobs_paused (HTTP 503) File jobs are paused. File jobs are paused for maintenance. What to do: Try again later. Tools that work on text input are not affected. Safe to retry unchanged: yes. ### job_timeout (HTTP 422) The job took too long. This job ran longer than its time limit and was stopped. What to do: Try a smaller or shorter file, or ask for smaller output. Safe to retry unchanged: no. ### job_out_of_memory (HTTP 422) The job ran out of memory. This job needed more memory than one job is allowed to use and was stopped. What to do: Try a smaller file. For images, reduce the pixel dimensions first. Safe to retry unchanged: no. ### job_output_too_large (HTTP 422) The result is too large. The files this job produced are larger than the output limit. What to do: Ask for smaller output, for example lower quality, smaller dimensions or fewer files. Safe to retry unchanged: no. ### worker_lost (HTTP 503) The worker stopped unexpectedly. The worker running this job stopped before it finished, and the job could not be completed after retrying. What to do: Run the tool again. If it fails the same way, the file itself may be triggering the problem. Safe to retry unchanged: yes. ### job_cancelled (HTTP 409) The job was cancelled. This job was cancelled before it finished. What to do: Run the tool again if you still need the result. Safe to retry unchanged: no. ### job_failed (HTTP 500) The job hit an unexpected problem. Something went wrong on our side while running this job. What to do: Try again. If it keeps happening, send the job ID to https://setget.net/feedback so we can find the cause. Safe to retry unchanged: yes. ### tool_unavailable (HTTP 503) This tool is not available on this server. This tool needs software that is not installed on the worker. What to do: Try again later, or look for a related tool at https://setget.net/tools. Safe to retry unchanged: no. ### invalid_json (HTTP 400) The request body is not valid JSON. The request body could not be read as a JSON object. What to do: Send a JSON object, for example {"input": {"text": "hello"}}, with Content-Type: application/json. Safe to retry unchanged: no. ### too_large (HTTP 413) The request is too large. The request is larger than the 1 MiB limit. What to do: Send a smaller request. For big files, upload the file and pass a reference instead of the content. Safe to retry unchanged: no. ### rate_limited (HTTP 429) Too many requests. Too many requests arrived from this network address in a short time. What to do: Slow down and try again after the number of seconds in the Retry-After header. Safe to retry unchanged: yes. ### daily_limit (HTTP 429) Daily write limit reached. You have reached the daily limit of 5,000 writes from this network address. What to do: Wait until tomorrow, or reduce how often you write. Reading is not affected. Safe to retry unchanged: no. ### not_found (HTTP 404) Nothing is stored there. Nothing is stored under that name. It may never have existed, or it may have expired. What to do: Check the name. Everything on SetGet expires, so create it again if it is gone. Safe to retry unchanged: no. ### forbidden (HTTP 403) This credential does not allow that. The credential you sent does not grant this operation. What to do: Use the write token for changes, or the password the key was protected with. Safe to retry unchanged: no. ### unavailable (HTTP 503) Storage is temporarily unavailable. The service could not reach its storage. What to do: Try again shortly. For a write, check the current state first so the change is not applied twice. Safe to retry unchanged: yes.