skills-doctor
Skill1 starsChanged 6 months agoMITView on GitHub
Use only for doctor, skills doctor, skills check, audit skills, scan skill directories, diagnose existing local AI agent skills, or generate a local skills health report. Do not use for creating, installing, finding, recommending, or learning to write skills.
4 other projects use a version of this.
Adopted by
Most recently taken up by
- oldwinter/skillstheir own version, first seen Nov 8, 2026
- wangyuchen5110/kimi-skills-doctortheir own version, first seen Nov 8, 2026
- thititongumpun/skillstheir own version, first seen Nov 2, 2026
- milad-ahmd/agent-skillstheir own version, first seen Nov 1, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- oldwinter/skills★ 11same name
- milad-ahmd/agent-skills★ 5same name
- thititongumpun/skills★ 1same name
- wangyuchen5110/kimi-skills-doctor★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Show its rules and text
We split this skill into the rules it gives an agent
16 rules. 0 are common, 10 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
10
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Use this skill to help the current AI agent inspect local AI agent skills and produce a best-practice health report.
- Distinctive: line 14
If the user asks to apply fixes, stop this review workflow and ask for explicit confirmation before making any file changes.
- Distinctive: line 19
If the user provides paths, inspect those paths.
- Distinctive: line 20
If not, inspect only existing paths from the known roots list below.
- Distinctive: line 21
Do not search the entire home directory.
- Distinctive: line 26
Prefer file search and directory listing tools over reading every file up front.
- Distinctive: line 28
Read only representative or suspicious files first.
- Very specific: line 29
Start with SKILL.md frontmatter and headings.
- Distinctive: line 31
Do not execute scripts from inspected skills unless the user explicitly asks and the command is safe.
- Very specific: line 34
Read references/review-checklist.md for the full audit rubric.
- Very specific: line 35
Read references/anti-patterns.md when explaining or rewriting a problematic skill.
- Very specific: line 36
Read references/report-template.md before producing a concise chat or Markdown report.
- Very specific: line 37
Read references/html-report-template.md when the user asks for an HTML, visual, shareable, or archiveable report.
- Distinctive: line 43
Check only paths that exist.
- Distinctive: line 53
Use the three-layer model when interpreting a skill:
- Very specific: line 63
Use references/review-checklist.md as the source of truth for qualitative judgment.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skills Doctor
Boundary
Use this skill to help the current AI agent inspect local AI agent skills and produce a best-practice health report.
This skill is read-only for inspected user skills. Do not edit, delete, move, install, uninstall, overwrite, or patch any user skill files, including SKILL.md, references/, scripts/, assets/, or agent configuration files. Only inspect files, summarize risks, and suggest next edits.
If the user asks to apply fixes, stop this review workflow and ask for explicit confirmation before making any file changes.
Workflow
- Confirm the target directories.
- If the user provides paths, inspect those paths.
- If not, inspect only existing paths from the known roots list below.
- Do not search the entire home directory.
- Build a lightweight inventory before deep reading skill bodies.
- Find skill directories that contain
SKILL.md. - Record each skill's path,
name,description, approximate size, and resource folders. - Prefer file search and directory listing tools over reading every file up front.
- Find skill directories that contain
- Read only representative or suspicious files first.
- Start with
SKILL.mdfrontmatter and headings. - Inspect
references/,scripts/, orassets/only when the issue depends on them. - Do not execute scripts from inspected skills unless the user explicitly asks and the command is safe.
- Start with
- Apply the review references:
- Read
references/review-checklist.mdfor the full audit rubric. - Read
references/anti-patterns.mdwhen explaining or rewriting a problematic skill. - Read
references/report-template.mdbefore producing a concise chat or Markdown report. - Read
references/html-report-template.mdwhen the user asks for an HTML, visual, shareable, or archiveable report.
- Read
- Summarize the highest-risk findings to the user with exact paths, evidence, impact, and concrete next edits. If generating HTML, write
skills-doctor-report.htmlunless the user requests another path.
Known Skill Roots
Check only paths that exist. This list is intentionally explicit so the agent does not scan the entire home directory.
- Universal:
.agents/skills,~/.agents/skills. - Project-level:
.codex/skills,.claude/skills,.cursor/skills. - Core global agents:
~/.codex/skills,~/.claude/skills,~/.cursor/skills. - Popular agent-specific roots:
~/.aider-desk/skills,~/.augment/skills,~/.bob/skills,~/.openclaw/skills,~/.codeartsdoer/skills,~/.codebuddy/skills,~/.codemaker/skills. - Additional known roots when present:
~/.amp/skills,~/.antigravity/skills,~/.cline/skills,~/.dexto/skills,~/.firebender/skills,~/.gemini/skills,~/.github-copilot/skills,~/.kimi/skills,~/.opencode/skills,~/.warp/skills.
Loading Model
Use the three-layer model when interpreting a skill:
- Index:
name + description, the trigger surface that may be visible before a skill is selected. - Load:
SKILL.md, the content loaded after a skill is selected. - Runtime:
references/,scripts/,assets/, and command output that may be loaded on demand.
Token counts are rough estimates for relative risk ranking, not billing numbers.
Review Dimensions
Use references/review-checklist.md as the source of truth for qualitative judgment. Treat deterministic observations, such as file size or repeated trigger terms, as leads rather than proof.
Output
Deliver either a concise chat/Markdown report or a local HTML report:
- scanned roots and skill count
- P1/P2/P3 finding count
- top risks grouped by Index, Load, and Runtime
- exact paths and evidence for each high-risk issue
- what the user should inspect or change next
For HTML reports, use references/html-report-template.md, keep the file self-contained, and tell the user the generated path.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

