setup-hooks

An opt-in setup command for documentation checks that run when code is committed, edited, or reviewed in CI. CI means automated checks run by the project after changes are proposed.

In plain words
What is it for?
Use it to install selected checks for Markdown style, documentation impact, links, references, and unsupported claims.
Why use it?
It catches documentation mistakes and reminds contributors to update docs when source changes affect them.

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/setup-hooks
Clone the repo
git clone --depth 1 https://github.com/EdwardAngert/docs-agent-plugin
Per session 41 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,960 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00041 $0.01960
Opus 5 $0.00020 $0.00980
Sonnet 5 $0.00008 $0.00392
Haiku 4.5 $0.00004 $0.00196

Measured 2d ago against content hash 5dc190d5ff9f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

setup-hooks 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 2d 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.

commands/setup-hooks.md · 92 lines

How it starts

The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Set Up Documentation Hooks

Install hooks that keep documentation in shape automatically. Hooks are opt-in: this command installs only what you choose, and it never turns anything on by itself. The optional argument ($ARGUMENTS) names a hook to skip the selection question.

Available Hooks

  • Git pre-commit hook (${CLAUDE_PLUGIN_ROOT}/assets/hooks/pre-commit): lints staged Markdown with the repo's configured linter and reminds you to run /docs-assist:update when you commit source changes without touching docs. Fails the commit only on lint errors. The reminder never blocks.
  • Claude Code post-edit lint (${CLAUDE_PLUGIN_ROOT}/assets/hooks/claude-code-hooks.json): a PostToolUse hook that lints a Markdown file right after Claude writes or edits it, so style issues surface in the session. Requires jq and npx.
  • CI docs-impact check (${CLAUDE_PLUGIN_ROOT}/assets/ci/docs-impact.mjs and ${CLAUDE_PLUGIN_ROOT}/assets/ci/github/docs-impact.yml): a deterministic detector that runs on every pull request and reports when a diff rides the change types that ripple into docs: moved or renamed docs, changed headings, changed code terms the docs mention, or a large source change with no docs touched. It costs no agent tokens; it tells reviewers when /docs-assist:update is worth running, and can be made blocking with DOCS_IMPACT_STRICT.
  • CI reference-registry check (${CLAUDE_PLUGIN_ROOT}/assets/ci/check-facts.mjs and ${CLAUDE_PLUGIN_ROOT}/assets/ci/github/check-facts.yml): a deterministic detector that runs on every pull request and verifies the mechanical parts of .docs-assist/reference.yml: every fact entry's source still contains the referenced identifier, and every pointer entry's target file and heading anchor still resolve. Only offer this when the project has a reference.yml with fact or pointer entries; skip it otherwise, since there's nothing for it to check. Can be made blocking with CHECK_FACTS_STRICT.
  • CI claim check (${CLAUDE_PLUGIN_ROOT}/assets/ci/check-claims.mjs and ${CLAUDE_PLUGIN_ROOT}/assets/ci/github/check-claims.yml): a deterministic detector that runs on every pull request, extracting checkable claims (CLI flags, config/env keys, function or class names, file paths, version requirements) from every doc and resolving each against the code with git grep. Unlike the reference-registry check, this needs no curated .docs-assist/reference.yml entries first; it scans the whole doc set cold. It only settles claims a lookup can settle: described-behavior and numeric claims still need claim-verification.md's agent trace. Can be made blocking with CHECK_CLAIMS_STRICT, though non-strict is the safer default since a "missing" result can also mean the claim's target is intentionally untracked (gitignored) rather than gone.

Read the full file on GitHub · 92 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. 2d ago First seen · 92 lines · 41 tokens per session scan A 5dc190d5ff9f

Subscribe to this mod's changes

setup-hooks is a command published in the GitHub repository EdwardAngert/docs-agent-plugin (5 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 41 tokens to every session and 1,960 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.