Utility tools for humans and agents
Small, deterministic tools — each with a human UI, REST API, OpenAPI schema, and MCP server entry. Free to use. Agent-native from day one.
Featured tools
For AI agents
- GET /openapi.json — OpenAPI 3.1 for all tools
- GET /llms.txt — llms.txt discovery surface
- POST /api/<slug> — REST API, JSON in/out
- MCP — same registry, stdio or HTTP
What is quickhelp.dev?
quickhelp.dev is a collection of small, focused developer tools that do exactly one thing well. Each tool is free to use — no account, no rate limit for normal use. You can use any tool through its web interface, call it directly as a REST API, discover it through an OpenAPI 3.1 schema, or register it as an MCP tool inside Claude, Cursor, or any MCP-compatible agent.
The tools are deterministic: given the same input, they always produce the same output. There are no black-box AI models deciding the result — just well-defined transformations like decoding a JWT, formatting JSON, converting an image, or removing a background. That predictability makes them reliable for automated workflows, CI pipelines, and AI agents that need a tool they can trust.
Three ways to use every tool
Visit the tool page, fill in the form, click Run. Results appear instantly and can be copied with one click. No sign-up, no tutorial.
Every tool has a POST /api/<slug> endpoint that accepts JSON and returns JSON. No API key required for standard use. Integrate it into any script, workflow, or application.
Add quickhelp.dev to Claude Desktop or any MCP-compatible agent. All tools are registered automatically — the agent can discover and call them without extra configuration.
Why deterministic tools?
AI agents work best with tools that are predictable. A tool that decodes a JWT should always return the same header and payload — not a summary, not a paraphrase, not a hallucination. quickhelp.dev tools are built with that contract in mind: strict JSON schemas for inputs and outputs, no side effects, no state between calls.
Every tool exposes its input and output shapes via OpenAPI 3.1, which means agents using the OpenAPI spec can validate inputs before calling and parse outputs without guesswork. The llms.txt file at the root of the domain gives language models a structured overview of what each tool does, how to call it, and what to expect back.
From the blog
In-depth guides on the formats and techniques these tools are built around.
- How to decode a JWT safely in 2026 (no signature verification)Learn how to read JWT header and payload without needing the signing secret, when this is safe, and when it isn't.
- PNG vs WebP vs AVIF: a 2026 benchmark using our converterWe converted 100 images across three formats and measured file size and quality. Here are the numbers.
- Why we run background removal in your browser, not on a serverServer-side AI costs money per call. WebAssembly doesn't. How we ship AI tools for free.
- JSON formatter benchmarks: which approach is fastest?Comparing JSON.stringify, streaming parsers, and tree-diffing across 1 KB to 1 MB payloads.