Generate man-page-style help for your skills
SaferSkills independently audited skill_AI_d (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Turn any system of skills into a self-documenting one.
You build a system of skills. It works well. Then nobody — including future-you — can remember which trigger does what. skill_AI_d solves this by producing a single [SYSTEM]_help.md file that Claude can serve on demand, directly in chat.
No code. No external tools. Just guided skill execution and Markdown output.
→ Full documentation: README.md → Source & install: github.com/playtosor/skill_AI_d
Requirements: Claude Desktop with filesystem MCP access.
Typical workflow:
cdc_help.md → produces a validated specification (CDC)help.md → produces [SYSTEM]_help.md[SYSTEM]_help.md includedupdate_help.md → delta appliedLoad the relevant skill from your skill_AI_d folder:
cdc_help.md — guided procedure to produce the specification before authoringhelp.md — structure convention + authoring of [SYSTEM]_help.mdupdate_help.md — delta detection and surgical update after system changes| Trigger | Skill | What it does |
|---|---|---|
UPDATE HELP? | update_help.md | Quick scan — signals what has changed, no file written |
UPDATE HELP! | update_help.md | Full delta analysis, validation, and update |
cdc_help.md is guided — no fixed trigger. Run it by reading the skill and following the steps.
Running skill_AI_d on a system produces one file: [SYSTEM]_help.md. Reference implementation: m_AI_l_help.md — the help file for the m_AI_l messaging system. <!-- END -->
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.