audit

A command for systematically reviewing documentation for quality, missing content, structure, and reader experience.

In plain words
What is it for?
Use it to audit whether documentation fits its audience, covers the needed tasks, and guides readers from entry to completion.
Why use it?
It helps find gaps and confusing journeys that formatting or spelling checks do not catch.

Command

Install

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.

agentmods
npx agentmods add commands/edwardangert/docs-agent-plugin/audit
Clone the repo
git clone --depth 1 https://github.com/EdwardAngert/docs-agent-plugin
Per session 11 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,688 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00011 $0.03688
Opus 5 $0.00005 $0.01844
Sonnet 5 $0.00002 $0.00738
Haiku 4.5 $0.00001 $0.00369

Measured yesterday against content hash 22ac8918eea4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

audit scanned grade A with 1 finding 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 yesterday.

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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

1. **No tooling at all, or a one-off scoped audit**: run `npx --yes markdown-link-check --quiet <files>` directly (add `--config .markdown-link-check.json` if present) rather than hand-rolling requests with `curl`/`WebFe
commands/audit.md · 221 lines

How it starts

The opening of the file, as written. The whole thing — 221 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Audit Documentation

Perform a systematic documentation audit on the specified path: $ARGUMENTS

Your Role

Focus on content and strategy, the things that require judgment:

  • Is this the right content type for the user's goal?
  • Does the information architecture match how users think?
  • Are there gaps in documentation coverage?
  • Does content make assumptions about prerequisite knowledge?
  • Is related information consolidated or scattered?
  • Are docs serving users or just describing features?
  • Does each doc's reader journey actually work? Outline the user stories each doc serves and walk them through it: arrival, entry, path, exit. See ${CLAUDE_PLUGIN_ROOT}/skills/docs-assist/reference/user-stories.md.

Leave mechanical checks to linters (Vale, markdownlint, cspell). This is an operation, not just a principle: see step 0 below. Do not reimplement MD022, MD032, MD040, or Vale's prose rules by reading files by hand; a linter finds every instance in seconds where a manual pass misses several per file. Running the linters clean is the floor, not the audit. It confirms the docs are well-formed; it says nothing about whether they're still true. Your primary value is the two things a linter cannot do: understanding the documentation holistically to serve users better, and tracing what the docs claim out to the code that should back it (see "Trace claims to the code" under Content Issues below). A pass that stops at a clean lint run has not audited the content yet.

Resolve .docs-assist/config.yml and style.md first if they exist, and audit against them. When they do not exist, do not stop to ask about conventions: the docs set's own internal consistency is the standard. Hold the set to the rigor of a full documentation team reviewing a solo writer's work: example values that drift between docs, the same concept under different terms, stale cross-references, and structural inconsistencies between sibling docs.

Audit Process

Match the depth of the audit to the target. A full documentation set, a single directory, a handful of changed files, and a diff are different jobs. Do not run whole-set steps against a few files.

Read the full file on GitHub · 221 lines

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.

  1. yesterday First seen · 221 lines · 11 tokens per session scan A 22ac8918eea4

Subscribe to this mod's changes

audit is a command published in the GitHub repository EdwardAngert/docs-agent-plugin (5 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 11 tokens to every session and 3,688 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.