Free tools · no signup

Tools for Claude skills

Eight working tools from our test bench. Most run in your browser; nothing is uploaded.

01 · Validator

SKILL.md validator

Paste your SKILL.md, get an instant score against the rubric we use in reviews. Catches the mechanics that sink roughly half the skills we test — vague trigger descriptions, missing frontmatter, thin bodies.

Check your SKILL.md

02 · Calculator

Skill token-cost calculator

Every installed skill's trigger description rides along in every request — the always-on tax. Count what your set costs per month. (Skill bodies load only on trigger, so this is the floor.)

Skills installed Avg description length, chars Requests per day Model input price, $/M tokens

—tokens / month

—$ / month, always-on tax

03 · Generator

SKILL.md generator

Answer four questions, get a correctly structured SKILL.md with a trigger description that actually fires. Built from the patterns of our highest-scoring skills.

Skill name (kebab-case) What it does, one line When should it trigger? (situations, one per line) Steps Claude should follow (one per line)

Your SKILL.md appears here…

04 · Converter

Rules ⇄ SKILL.md converter — both directions

Works both ways: wrap your .cursorrules / AGENTS.md / GEMINI.md / copilot-instructions.md as a Claude skill — or unwrap any SKILL.md into the format Cursor, OpenAI Codex, Gemini CLI or GitHub Copilot expects.

Direction Rules format Skill name (kebab-case) When should Claude apply these rules? Paste your rules

Your SKILL.md appears here…

05 · Picker

Skill picker

Two answers — three tested skills matched from the catalog by real scores, plus the pre-configured pack for your role if you'd rather skip the setup.

What do you do? Biggest pain right now Or describe your task in your own words (optional — overrides the dropdowns)

06 · For authors

Tested badge finder

Your skill passed our independent testing? Pick it — get the embeddable badge that links back to its public test report.

Your skill

Not listed yet? Submit your skill — testing is free.

07 · MCP server

Ask your agent, not a search box

Install this and Claude Code checks a skill's verdict before it installs anything — including the ones we found broken.

claude mcp add skillproof -- npx -y skillproof-mcp

Two tools: find_skill ("is there a tested skill for X?") and check_skill ("someone recommended Y — does it work?"). Each answer carries the verdict, the /10 score, what our test actually found, and the install command. Works in any MCP client — Claude Code, Claude Desktop, Cursor, Windsurf, Zed. No key, no account.

Source: github.com/Skillproofdev/skillproof-mcp. Building your own tooling? The catalog is open: /api/skills.json and /llms-full.txt (CC BY — cite skillproof.dev).

08 · Install our skill

Let Claude pick your skills for you

Prefer a plain skill over an MCP server? Install find-tested-skills and just ask Claude Code "what skills should I use for this project?" — it reads your repo, queries this catalog, and hands back the 2–4 best-scoring tested matches with install commands. Same data as the MCP server, zero setup: it's one SKILL.md.

git clone https://github.com/Skillproofdev/find-tested-skills.git
cp -r find-tested-skills ~/.claude/skills/find-tested-skills

Then, in Claude Code: "find me a tested skill for writing SQL migrations" or "which skills should I install here?" It never recommends a skill that scored fails — and tells you when a popular one did. CC BY 4.0; reads the open /api/skills.json.

Can't find what you need?

Request a skill — we'll test or build it

Tell us the job. Within a few days you get back a link to a tested skill that already does it, a SKILL.md you can build from, or a ready-to-use prompt. Popular asks get built for the catalog.

★ 9.6/10 × 3

The free starter pack

3 skills with our highest test scores plus the install checklist — the setup we'd put on a fresh machine. Free, by email.