agentleFS
DGouronDGouron/review-flow/.claude/skills/audit-docs/SKILL.md

audit-docs

Skill42 starsChanged 6 months agoMITView on GitHub

Audit documentation for duplication, staleness, language issues, and verbosity. Read-only — produces an actionable report.

17 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use diagnose.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    diagnose

    Use to systematically diagnose why scenarios fail — identifies root causes by analyzing agent traces, codebase, and behavioral patterns

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 4 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 16 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read .claude/roles/documentalist.md — adopt this profile and follow all its rules.

    line 17
  2. Distinctive:

    Title overlap: if two docs share > 40% of their H2/H3, flag them

    line 45
  3. Very specific:

    Duplicate markers: look for concept definitions (e.g., THREAD_RESOLVE, phase names) that appear in multiple files

    line 47
  4. Very specific:

    Check last-updated in the frontmatter (if present)

    line 57
  5. Very specific:

    If the source is more recent than last-updated, flag as potentially stale

    line 62
  6. Distinctive:

    Verify referenced paths in the doc — whether they still exist:

    line 63
  7. Distinctive:

    Look for references to concepts/APIs that no longer exist

    line 68
  8. Distinctive:

    Rule: all documentation must be in English.

    line 77

“Common” is measured: kept by at least 6 in 10 of the 16 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Audit Documentation

Activation

This skill activates for:

  • Checking documentation quality
  • Detecting duplicates, stale docs, inconsistencies
  • /audit-docs

Persona

Read .claude/roles/documentalist.md — adopt this profile and follow all its rules.

Mode

READ-ONLY. This skill analyzes and reports. It does not modify any file.

Workflow — 5 Audit Passes

Pass 1: Inventory

Glob docs/**/*.md
Glob templates/**/*.md

For each file, extract:

  • Title (H1 or frontmatter title)
  • Frontmatter present? (yes/no)
  • Headings H2/H3
  • Detected language
  • Approximate word count

Produce an inventory table.

Pass 2: Duplicate Detection

Compare H2/H3 headings across all file pairs:

  1. Title overlap: if two docs share > 40% of their H2/H3, flag them
  2. Similar content: for flagged pairs, compare the first 100 words of each section
  3. Duplicate markers: look for concept definitions (e.g., THREAD_RESOLVE, phase names) that appear in multiple files

Known issues to detect:

  • DEPLOYMENT.md vs deployment/README.md — nearly identical
  • PROJECT_CONFIG.md vs CONFIG-REFERENCE.md — project config overlap
  • Markers documented in 5+ files
  • Troubleshooting duplicated in 3+ files

Pass 3: Staleness Detection

  1. Check last-updated in the frontmatter (if present)
  2. For each related file in the frontmatter:
    git log -1 --format=%ci <related-file>
    If the source is more recent than last-updated, flag as potentially stale
  3. Verify referenced paths in the doc — whether they still exist:
    # For each src/... path mentioned in the doc
    ls <path>
  4. Look for references to concepts/APIs that no longer exist

Pass 4: Language Consistency

Detect French in each file. Indicators:

  • Frequent words: "les", "des", "une", "est", "dans", "pour", "avec", "cette", "qui", "sur"
  • Entire sentences in French
  • Titles in French

Rule: all documentation must be in English.

Pass 5: Verbosity Analysis

ThresholdAction
Section > 500 wordsSuggest condensation or split
Document > 800 wordsSuggest split into multiple docs
Filler phrases detectedList with location

Filler phrases to detect:

  • "it should be noted that", "it is important to mention"
  • "as you can see", "as mentioned above/below"
  • "in order to" (use "to"), "due to the fact that" (use "because")
  • "at the end of the day", "going forward"

Report Format

# Documentation Audit Report

## Summary

| Metric | Count |
|--------|-------|
| Total docs scanned | X |
| Duplication issues | X |
| Stale docs | X |
| Language issues | X |
| Verbosity issues | X |
| Missing frontmatter | X |

## Duplication Issues

### DUPL-001: [Doc A] <-> [Doc B]
- **Topic**: [overlapping topic]
- **Severity**: high | medium | low
- **Overlapping sections**: [list]
- **Action**: Consolidate into [file], link from [other file]

## Stale Documentation

### STALE-001: [Doc]
- **Last updated**: [date or "no frontmatter"]
- **Source changed**: [file] on [date]
- **Action**: Update [specific sections]

## Language Issues

### LANG-001: [Doc]
- **Language**: French (should be English)
- **Action**: Translate to English

## Verbosity Issues

### VERB-001: [Doc] — [Section]
- **Word count**: X (target: < 300)
- **Fillers found**: [list]
- **Action**: Condense

## Recommended Actions (Priority Order)

1. [Highest impact action]
2. [...]

After the Audit

Recommend the appropriate skills:

  • Duplicates: /update-docs to consolidate
  • New doc needed: /create-doc
  • Missing/stale index: /docs-index

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.