Guide
Skill quality leaderboard: what a 90+ skill looks like
Every free ProSkills skill, ranked by what the scanner says about the files it ships. Below it, three made-up samples scored the same way, and a breakdown of what separates a 90+ skill from the rest.
The free skills, ranked
Scores were computed from the shipped files when this page was built. The rows link to each skill.
| # | Skill | Scan | Grade | Loads on trigger |
|---|---|---|---|---|
| 1 | AGENTS.md Architect | 100/100 | A | 1,702 tokens |
| 2 | Analytics SQL | 100/100 | A | 1,812 tokens |
| 3 | Character Bible | 100/100 | A | 1,189 tokens |
| 4 | CRUD Scaffold | 100/100 | A | 1,038 tokens |
| 5 | Dockerfile Doctor | 100/100 | A | 1,737 tokens |
| 6 | Grill Me | 100/100 | A | 1,075 tokens |
| 7 | Hook Lab | 100/100 | A | 971 tokens |
| 8 | Humanizer | 100/100 | A | 1,024 tokens |
| 9 | Hyperframes Starter | 100/100 | A | 1,055 tokens |
| 10 | PR Describer | 100/100 | A | 1,207 tokens |
| 11 | Prospect Brief | 100/100 | A | 952 tokens |
| 12 | Taste Core | 100/100 | A | 951 tokens |
| 13 | Wardrobe & Sets | 100/100 | A | 945 tokens |
The 49 paid skills are scanned the same way before they ship. Their scores run from 100 to 100, and each is shown on its own page.
Three synthetic samples
| Sample | Scan | Grade | Why |
|---|---|---|---|
| Sample A, written thin | 69/100 | D | eff/body-thin, eff/description-thin, eff/description-trigger |
| Sample B, half finished | 81/100 | B | eff/body-thin, clarity/examples, clarity/output |
| Sample C, full structure | 100/100 | A | No findings |
Spec 25 · Efficiency 3 · Clarity 16 · Security 25 (each out of 25)
Always loaded 10 tokens · on trigger 27 · on demand 0
eff/body-thinwarnThe body is very thin. A skill should carry knowledge the agent doesn't already have.eff/description-thinwarndescription is too short to trigger reliably. Say what it does and when to use it.eff/description-triggerwarndescription doesn't say *when* to use the skill ("Use when …"). Agents decide from this line alone.clarity/exampleswarnNo examples or code blocks. One concrete input → output example improves consistency more than extra rules.- and 1 more.
Spec 25 · Efficiency 15 · Clarity 16 · Security 25 (each out of 25)
Always loaded 29 tokens · on trigger 58 · on demand 0
eff/body-thinwarnThe body is very thin. A skill should carry knowledge the agent doesn't already have.clarity/exampleswarnNo examples or code blocks. One concrete input → output example improves consistency more than extra rules.clarity/outputwarnDoesn't say what the result should look like. Add an Output section.
Spec 25 · Efficiency 25 · Clarity 25 · Security 25 (each out of 25)
Always loaded 81 tokens · on trigger 549 · on demand 112
No findings.
Anatomy of a 90+ skill
Sample C above has no findings. Here is what it does, by check. The points are the scanner's: 25 for spec, 25 for efficiency, 25 for clarity and 25 for security, each out of 25.
| Check | What the skill does | Rules that cost points when missing |
|---|---|---|
| Spec | Valid frontmatter, a lowercase name that matches its folder, a description under 1,024 characters, links that resolve. | spec/frontmatter, spec/name-format, spec/broken-ref |
| Efficiency | A description that says when to use it, a body under 5,000 tokens, detail in a reference file that the body names, enough specifics to beat a plain prompt. | eff/description-trigger, eff/body-tokens, eff/body-thin |
| Clarity | Numbered steps, at least one worked example, a stated output, no unfinished markers. | clarity/examples, clarity/output, clarity/unfinished |
| Security | Nothing that overrides instructions, hides text, runs remote code or touches credentials. | sec/prompt-injection, sec/hidden-unicode, sec/remote-exec |
The full rule list is in the validator guide, and the format itself in the format reference.
Get your own skill to 90
Run it through the scanner, fix the findings in order, and rescan. Or have an agent do the writing: Skill Author ($12) is built to produce a skill that scores 90 or more, and it is scanned by the same rules itself (100/100). Start from how to write a skill if you would rather do it by hand.
Questions
- How is the score calculated?
- Four checks of 25 points each: spec, efficiency, clarity and security. Findings take points off the check they belong to. Any critical finding caps the grade at F whatever the total.
- Does a high score mean a skill is good?
- It means the skill is well formed, cheap to load, clear and free of known risky patterns. It does not measure whether the advice is right for your work. Read the skill and try it.
- Why is a skill under 90?
- Common reasons are a body that is thin on specifics, no worked example, no stated output, or a missing trigger phrase in the description. Each finding names its rule, so the fix is listed.
- Are the samples real skills?
- No. The three samples are written for this page, named so they cannot be mistaken for any real skill, and scored by the same scanner. The leaderboard rows are the real free skills in the ProSkills catalog.
Keep reading
- Skill ScannerScore any skill in your browser. Nothing is uploaded.
- SKILL.md format referenceEvery field, its limit, the folder layout and the errors people hit.
- SKILL.md examplesReal skills, annotated, each with its scan score.
- How to write a Claude skillSeven steps, a weak and a strong description, scored side by side.
- All guidesEvery guide, grouped by what you are trying to do.
Paste a SKILL.md or drop a folder. It is scored in your browser and nothing is uploaded.
Want a skill that already passes? Start with a free one.