Free skill description tester
Claude Code decides whether to load a skill almost entirely from the description (and optional when_to_use). Paste yours below, add a few real requests you'd type, and this page scores the match with honest heuristics — word overlap, formula, length, voice and YAML safety. It is not Claude's real matcher; the real test is asking for the task in a fresh session.
Everything runs in your browser. Nothing you paste is sent anywhere.
Results
What the checks are based on
- Formula: a description that triggers reliably starts with a verb (what it does / the output) and includes "when to use" wording that matches how people ask.
- Length: under ~8 words is usually too vague; over 1,024 characters hits the Agent Skills / claude.ai cap. In Claude Code the listing cuts
description+when_to_useat 1,536 characters, so put the key use case first. - Third person: write "Drafts a weekly update…", not "I draft…" or "You can use this to…".
- No XML tags and YAML-safe quoting: an unquoted value with
:in the middle can break frontmatter parsing, which silently empties the description. - Word overlap: a light stem + stop-word filter comparing your asks to the description. Useful as a hint for missing trigger words; not how Claude actually decides.
Sources: Claude Code skills docs and Agent Skills best practices.
Next steps: build a fresh file with the SKILL.md builder, lint the whole frontmatter with the SKILL.md validator, or read how to write a description that triggers and the 10-point skill-not-triggering checklist.
Made by Solo Stack Kits. If you'd rather start from skills that already pass, the free 3-skill sample has ready-made skills for proposals and client updates, and the Solo Consultant Kit adds the full set.