Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Jerry0022/dotclaude --skill plugin-auditgit clone --depth 1 https://github.com/Jerry0022/dotclaudeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jerry0022/dotclaude/plugin-audit)<a href="https://agentmods.dev/skills/jerry0022/dotclaude/plugin-audit"><img src="https://agentmods.dev/badge/skills/jerry0022/dotclaude/plugin-audit/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/jerry0022/dotclaude/plugin-audit"><img src="https://agentmods.dev/badge/skills/jerry0022/dotclaude/plugin-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00126 | $0.01638 |
| Opus 5.5 | $0.00050 | $0.00655 |
| Sonnet 5.5 | $0.00025 | $0.00328 |
| Haiku 4.5 | $0.00013 | $0.00164 |
Grade A, and why
plugin-audit scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
plugin-audit
Audit of the plugin source, in five steps. Steps 1 and 5 are plain node (0 model
tokens); steps 2 and 3 are one subagent each; step 4 is plain node plus the
auto-concept hand-off. Rules produce findings — agents only prioritise and phrase.
State lives in <main repo>/.claude/plugin-audit/ (git-ignored, shared by all
worktrees): last.json, history.json, runs/<version>-<timestamp>/,
report-pending.json. <dir> below is the run directory.
Pending report first. If <state>/report-pending.json exists at the start of a
session (a ship-triggered run finished without a live session), skip step 1: its
dir holds results.json — continue with step 2 (skip a step whose output file
already exists in <dir>).
Step 1 — Collect (deterministic)
node .claude/skills/plugin-audit/scripts/audit.js --out <dir> [--skip-tests] [--days N] [--trigger manual|ship]
Without --out a new runs/ directory is created. From a ship, run it with the
Bash tool's run_in_background: true and carry on. It prints one summary line and
writes <dir>/results.json (footprint, mining, benefit, integrity[12], hygiene[8],
net). --skip-tests skips npm test + npm run lint (a full audit runs them).
Exit 0 unless the arguments are invalid.
Step 2 — Claude-Code currency (one devops:research agent, model: sonnet)
Spawn ONE agent. Brief it with: the Claude Code version recorded in the last audit
(history.json → last entry ccVersion; none → the last 60 days), the task "read the
Claude Code changelog since then and compare it with (a) the hook events and fields
the plugin uses (plugins/devops/hooks/hooks.json, hooks/lib/*.js transcript and
hook-input parsing) and (b) custom workarounds that a new Claude Code feature may now
make redundant (grep for workaround, hack, until Claude Code in hooks and
deep-knowledge)". It returns {ccVersion, findings:[{msg, file?, line?, kind: "breaking"|"opportunity"|"info"}]}.
Merge into <dir>/results.json → integrity check cc-currency: status red when a
breaking finding exists, yellow for opportunity, green when only info/none;
findings as returned; metrics: {ccVersion}. Agent failed or no network → leave
the check skipped and add the reason as a finding.
What ships with it
19 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- scripts/__fixtures__/transcripts/proj-a/agent-ignored.jsonl 314 B
- scripts/__fixtures__/transcripts/proj-a/s1.jsonl 2.0 KB
- scripts/audit.js 8.4 KB runs code
- scripts/audit.test.js 1.9 KB runs code
- scripts/dispatch-bench.js 6.6 KB runs code
- scripts/hook-audit.js 12 KB runs code
- scripts/integrity.test.js 3.3 KB runs code
- scripts/lib/footprint.js 9.5 KB runs code
- scripts/lib/hygiene.js 13 KB runs code
- scripts/lib/integrity.js 24 KB runs code
- scripts/lib/mining.js 18 KB runs code
- scripts/lib/state.js 2.8 KB runs code
- scripts/lib/trigger.js 2.1 KB runs code
- scripts/lib/util.js 3.3 KB runs code
- scripts/mining.test.js 6.1 KB runs code
- scripts/report.js 16 KB runs code
- scripts/report.test.js 5.5 KB runs code
- scripts/state.test.js 2.2 KB runs code
- scripts/trigger.test.js 2.2 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago Changed · +15 lines 9665767c1337
- 16d ago First seen · 113 lines · 126 tokens per session scan A d0e758ae8f36
plugin-audit is a skill published in the GitHub repository Jerry0022/dotclaude (6 stars, last pushed 11d ago), licensed MIT. It adds 126 tokens to every session and 1,638 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-21.
plugin-audit compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| plugin-auditJerry0022/dotclaude | Skill | 6 | 11d ago | MIT | A | 126 | Claude Code |
| autoreviewopenclaw/clawhub | Skill | 9.5k | yesterday | MIT | A | 17 | Claude Code, Codex |
| agent-teams-simplify-and-hardenLeoYeAI/openclaw-master-skills | Skill | 2.2k | 6d ago | MIT | A | 115 | Claude Code, Codex |
| omh-code-reviewrlaope/oh-my-hermes | Skill | 5.2k | yesterday | MIT | A | 48 | Claude Code, Codex |
| revdiff-planumputun/revdiff | Skill | 945 | yesterday | MIT | A | 84 | Claude Code |
| explore-unknownsdzhng/skills | Skill | 1.4k | 2d ago | MIT | A | 162 | Claude Code, Codex |
| re0-mergeLilMGenius/paperthin | Skill | 1.4k | 4d ago | MIT | A | 70 | Claude Code |