Borrowing it
Nothing to install: this file belongs to fulminate-io/knowledge-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/fulminate-io/knowledge-mcp/main/.claude/skills/audit-plan/SKILL.mdgit clone --depth 1 https://github.com/fulminate-io/knowledge-mcpWrote 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/fulminate-io/knowledge-mcp/audit-plan)<a href="https://agentmods.dev/skills/fulminate-io/knowledge-mcp/audit-plan"><img src="https://agentmods.dev/badge/skills/fulminate-io/knowledge-mcp/audit-plan.svg" alt="Measured on agentmods" 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 | $0.00059 | $0.01338 |
| Opus 5 | $0.00030 | $0.00669 |
| Sonnet 5 | $0.00012 | $0.00268 |
| Haiku 4.5 | $0.00006 | $0.00134 |
Grade A, and why
audit-plan 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 3d 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 — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AUDIT-PLAN — the adversarial audit procedure detail
Intent-fidelity audit
For every rule the plan asserts, locate the ORIGINAL statement (ticket quote, recorded decision) and diff semantically: duty-holder, cost-bearer, prevent-vs-compensate, absolute-vs-best-effort. Drift is T1/T2 — it invalidates every downstream artifact. A mechanism that only executes in a state the rule forbids is evidence the premise twisted upstream — the mechanism's existence is the finding even when its implementation is flawless. Vocabulary sweeps cover inflections and verb forms; a suspiciously clean census widens the pattern.
Requirement → criterion coverage
Walk In Scope as a checklist; every required behavior maps to a criterion that would FAIL if absent; decompose compounds ("X and Y" is two behaviors — the silent-omission path ships X, drops Y, looks complete). A specified behavior with no failing-when-absent catcher is T2.
Check each rule against the number beside it
For every RULE a plan states, check the FIGURE reported beside it is reachable under the rule AS WRITTEN. The recurring defect: the author probed one way and wrote the spec another, so the measurement was taken under corrected behavior while the prose describes the uncorrected one. The tell: a figure that cannot be derived from the rule printed next to it, or one BELOW an unfixed baseline. On a heavily revised plan this is the single highest-yield check.
Synthesize the post-edit artifact
Build the artifact the plan's own mandated text WOULD produce, then run every gate against it. It answers in one pass whether any criterion is unsatisfiable by the text its own step prescribes, and whether two criteria contradict each other. Cheap whenever the edits are enumerable. Build the faithful variant too (see the probe-a-gate rulebook).
Walk criteria to their artifacts
For every criterion, identify the EXACT artifact it reads and verify against THAT artifact, never the paragraph describing it. Two directed probes per plan: CROSS-NODE CONTRACT DIFF (where two nodes produce and consume the same field, diff exact spellings — two gates demanding opposite spellings of one field is a proven finding class) and SCOPE CHECK (what scope does each rule sentence bind, and does any gate apply it at a different scope?).
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.
- 3d ago First seen · 117 lines · 59 tokens per session scan A 5aa7e96a6878
audit-plan is a skill published in the GitHub repository fulminate-io/knowledge-mcp (1 stars, last pushed 2d ago), licensed Apache-2.0. It adds 59 tokens to every session and 1,338 once invoked, about $0.0003 per session on Opus 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 2026-09-04.
Other skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
rework-rate
Measure and interpret PR rework rate — the emerging 5th DORA metric.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
agent-teams-simplify-and-harden
Implementation + audit loop using parallel agent teams with structured simplify, harden, and document passes. Spawns implementation agents to do the work, then audit agents to find complexity, security gaps, and spec deviations, then loops until code compiles cleanly, all tests pass, and auditors find zero issues or…