Featured slot
Advertise to 36,000+ monthly visitors
Sponsored placement
SetupClaw: done-for-you OpenClaw for founders & exec teams
Sponsored
One API to scrape, enrich, and extract the internet.
Sponsored placement
CLN.Work — Stop prompting, start hiring AI employees
Affiliate
Launch your always-on OpenClaw agent on Hostinger, up to 70% off
Affiliate
Get a Hermes agent running 24/7 on Hostinger, up to 70% off
Free tool
SKILL.md Scorer
Paste your SKILL.md and get a 0-100 quality score with a letter grade and specific fixes — a name, a description that says when to use it, structure, examples, and a check for leaked secrets.
Scored in your browser. Your file never leaves this page.
Optional
Get the skill-authoring checklist + weekly digest
One useful issue a week on writing skills that agents actually route to, plus the marketplace-ready checklist. No spam, unsubscribe anytime.
How the SKILL.md score works
A SKILL.md is only useful if an agent can find it, decide when to use it, and trust it. This scorer checks the file against the signals that actually drive that — deterministically, so the same file always earns the same grade. It rewards a real name, a description that says WHEN to use the skill (the single highest-weighted check, because that is what routing depends on), a body with clear structure and length, concrete examples or usage steps, and it deducts hard if it spots a hardcoded secret. You get a score out of 100, an A–D grade using the same bands as our MCP scoring methodology, and a prioritized fix list. If you want a strict spec conformance pass instead, use the SKILL.md validator.
Everything runs in your browser — paste a full SKILL.md or just its YAML frontmatter and body, and nothing is uploaded. Once you like your score, submit the skill to the skills directory so builders and agents can discover it.
Frequently asked questions
What is the most important thing in a SKILL.md?
The description — specifically, whether it says when to use the skill. Agents read descriptions to decide which skill to load, so a vague description means your skill never gets picked, no matter how good the body is. That is why this check carries the most weight.
Why did my skill get flagged for a secret?
The scorer looks for patterns that match real credentials — OpenAI keys, GitHub tokens, AWS keys, bearer tokens, and assigned password values. Placeholders like ${API_KEY} or YOUR_API_KEY are ignored. If a real key was flagged, remove it, use an environment variable, and rotate the key immediately.
Scorer vs validator — which should I run?
The validator checks strict conformance to the SKILL.md spec (required fields, kebab-case, frontmatter syntax). The scorer is a quality lens: it grades how well the file is written to be discovered and trusted. Run the validator to pass the spec, then the scorer to stand out.
Does this tool send my data anywhere?
No. Scoring happens entirely in your browser. Your SKILL.md is never sent to a server unless you choose to enter your email for the digest.