String Utility Belt

Stop juggling online converters. Build your own text transformation pipelines. String Utility Belt is a free, open-source toolkit with 246 utilities, visual pipelines, per-step previews, and local processing. Use it in your browser, VS Code, terminal or AI workflows through MCP.

String Utility Belt (stringutilitybelt.com) is a free, open-source toolkit for transforming text. Instead of pasting text into one single-purpose converter after another, you build a pipeline: paste your input, add steps, and watch the result update as you type. Each step applies one utility to the output of the step before it. Turn on previews to see each step's output under its card and find exactly where a chain goes wrong.

It is built for developers, DevOps and support engineers, data analysts and anyone who reshapes text often: decoding tokens and log lines, converting between data formats, cleaning up pasted text or generating test data.

248 utilities so far. Encoding and decoding (Base64, hex, URL, HTML entities, Punycode and more), hashing (MD5, SHA-2, SHA-3, BLAKE, HMAC, bcrypt, Argon2), ciphers, compression, data formats (JSON, YAML, TOML, CSV, XML, INI, .env), line and text operations, analysis, generators, web and developer helpers, numbers, dates and colours. Values convert between text, raw bytes and JSON automatically, so a gzip step can feed a Base64 step with no glue. Magic mode detects what you pasted (Base64, a JWT, gzip, JSON and more) and suggests the step that decodes it.

More than a list of steps. • Branch steps send the same input down parallel lanes and merge the results. • Macros group a run of steps into one reusable step you can save to your library. • Run on each steps apply a sub-chain to every line, every delimited piece, every JSON array element or every JSON object value, then put the results back in place: decode one JWT per log line, or Base64-decode every value of a Kubernetes Secret. • Any step can have a run condition and its own error policy.

Share and reuse. A share link carries the whole pipeline, and your input only if you tick the box. The library keeps named pipelines and macros, with JSON import and export. 23 ready-made presets cover tasks such as decoding a Kubernetes Secret, cleaning text pasted from Word and turning nested JSON into CSV.

Your text stays on your device. Every transformation runs in your browser, and your input is never uploaded to be processed. There are no accounts and no cookies, and a share link keeps its contents after the #, which browsers don't send to the server. The site installs as an app and works offline.

Accurate by design. Every worked example on the site runs as an automated test against the real code, so the output you see is what the tool produces.

The same engine, outside the browser. • subelt, a command-line tool: steps as arguments, stdin to stdout. • An HTTP API (POST /api/run) for running pipelines from your own code. • An MCP server that lets AI agents such as Claude Code, Claude Desktop and Cursor call the utilities and pipelines as tools, so a hash or a decode comes out exact rather than guessed. • A Chrome extension that runs a favourite utility or saved pipeline on selected text from the right-click menu. • A VS Code extension, also on Open VSX for Cursor and VSCodium, that transforms the selection or replays a saved pipeline. • A Node.js library, @string-utility-belt/core.

From the listing. The maker can edit it.