refresh-tool-inventory — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited refresh-tool-inventory (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.
Regenerate the tool/prompt/resource tables in AGENTS.md and README.md from the source of truth (the Go packages in internal/).
Parse $ARGUMENTS:
--apply — write changes. Default is dry-run (print diff only). grep -nE 'mcp\.(AddTool|AddPrompt|AddResource)' \
internal/tools internal/prompts internal/resources -rCapture kind (tool/prompt/resource), name, handler, file:line.
definition to pull Description: / Title: fields. If absent, record description as <missing> and flag in the report.
AGENTS.md andREADME.md using these sentinel markers (REQUIRED — abort if missing):
<!-- inventory:tools:start --> ... <!-- inventory:tools:end -->
<!-- inventory:prompts:start --> ... <!-- inventory:prompts:end -->
<!-- inventory:resources:start --> ... <!-- inventory:resources:end -->If any sentinel is missing, abort with: ERROR: sentinel <name> missing in <file> — add markers before re-running. Never guess table boundaries.
the regenerated tables.
each matching sentinel pair. Leave everything outside the sentinels untouched.
Inventory diff (tools | prompts | resources):
Added: <name> (<kind>) — <file:line>
Removed: <name> (<kind>)
Renamed: <old> → <new>
Changed: <name> — description updated
Files that would be updated: AGENTS.md, README.md
Mode: dry-run | applyOn dry-run, print the diff and exit. On --apply, also write and print the final file paths.
Prefer promoting the extraction step to a deterministic shell or Go script under scripts/ once the format stabilises — this skill should then shrink to "invoke the script, preview, apply" (per the project's deterministic hierarchy rule).
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.