What is claudemd-auditor?

A review agent for an existing CLAUDE.md file, which is a project instruction file for Claude Code. It checks the file against a quality rubric and verifies whether its referenced paths, symbols, versions, and commands are still accurate.

Use it to establish a baseline score and find stale or incorrect instructions in one CLAUDE.md file. It also checks whether confirmed stale references appear in other CLAUDE.md files.

Project instructions can become outdated as files, scripts, and dependencies change. This audit separates confirmed problems from references that only look suspicious.

claudemd-auditor — Agent for Claude Code by Sage-Bionetworks

Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code

/plugin marketplace add Sage-Bionetworks/agent-skills

Claude Code

/plugin install agent-skills

Made for: Claude Code.

Or install agent-skills, the plugin that ships this one along with the rest of its 7 skills, 5 agents.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/agents/sage-bionetworks/agent-skills/claudemd-auditor/github.svg)](https://agentmods.dev/agents/sage-bionetworks/agent-skills/claudemd-auditor)

Your own site

<a href="https://agentmods.dev/agents/sage-bionetworks/agent-skills/claudemd-auditor"><img src="https://agentmods.dev/badge/agents/sage-bionetworks/agent-skills/claudemd-auditor/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.

Your own site · 80×15

<a href="https://agentmods.dev/agents/sage-bionetworks/agent-skills/claudemd-auditor"><img src="https://agentmods.dev/badge/agents/sage-bionetworks/agent-skills/claudemd-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 38 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 530 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00038	$0.00530
Opus 5	$0.00019	$0.00265
Sonnet 5	$0.00008	$0.00106
Haiku 4.5	$0.00004	$0.00053

Measured 12d ago against content hash 3feeffa3a95c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

claudemd-auditor 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.

agents/claudemd-auditor.md · 38 lines

What it actually says

CLAUDE.md Auditor

You support the evolve-claudemd skill. For a single CLAUDE.md you produce a baseline quality score and a list of confirmed accuracy issues. You do the mechanical verification via the skill's script so you only reason about the suspect subset — you don't Grep/Glob every line yourself.

Inputs

claude_md_path — the file to audit

Workflow

Run bash "${CLAUDE_PLUGIN_ROOT}"/skills/evolve-claudemd/scripts/verify-references.sh <claude_md_path>. It returns PATH_MISSING / SYMBOL_ABSENT (suspect) and PATH_OK / SYMBOL_OK (likely fine).
For each SUSPECT item, confirm it's genuinely stale (a basename-only reference or glob may resolve elsewhere — check before flagging). Discard false positives.
- When a reference IS confirmed stale (a deleted file/symbol), grep -rn the same token across all CLAUDE.md files in the repo — the same stale name (e.g. a removed config file) is usually cited in more than one file. Report every hit, noting it's a cross-file staleness.
Verify things the script leaves to judgment:
- Version numbers vs the actual build file (pom.xml/package.json) — read and compare.
- Commands vs package.json scripts / Makefile / CI — present verbatim?
- Architecture/Related-Systems prose still matches reality?
Score the file against "${CLAUDE_PLUGIN_ROOT}"/skills/evolve-claudemd/references/scoring-rubric.md (read it once). Record letter grade.

Output (return ONLY this)

## Audit — <claude_md_path>

Baseline score: <NN>/100 (<letter>)

### Confirmed accuracy issues
- Line <n>: <what's wrong> → <correction>

### Version/command mismatches
- Line <n>: "<current>" → "<actual>" (source: <build file>)

List only CONFIRMED issues (script false-positives removed). Keep under ~30 lines. Do not paste the whole CLAUDE.md or the raw script output.

Changes

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 First seen · 38 lines · 38 tokens per session scan A 3feeffa3a95c

Subscribe to this mod's changes

claudemd-auditor is an agent published in the GitHub repository Sage-Bionetworks/agent-skills (5 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 38 tokens to every session and 530 once invoked, about $0.0002 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 2026-08-31.

Code review

cpp-reviewer

Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.

affaan-m/ECC · 41 tokens

reviewer

Read-only reviewer for an SDD implementation — checks that the change satisfies the acceptance criteria it claims (stage 1) and meets quality/convention/edge-case bars (stage 2). Use after a task (or the whole feature) reaches GREEN, before it's considered done. It reads the diff and the upstream artifacts and reports…

genkovich/sdd · 81 tokens

atomic-auditor

Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…

damusix/atomic-claude · 169 tokens

claudemd-auditor