skills/doc-health-audit/SKILL.md
Orchestrate a full documentation health audit across five dimensions: structural health, semantic accuracy, topic completeness, prose quality, and information architecture. Runs each phase in dependency order with phase gating. This skill should be used for pre-release audits, periodic health checks, or comprehensive documentation assessments.
npx skillsauth add nickcrew/claude-cortex doc-health-auditInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Orchestrate a comprehensive documentation assessment across five dimensions. Each phase builds on the previous — structural issues are found before accuracy is checked, accuracy before completeness, and so on.
This skill does not duplicate any phase logic. It defines the run order, phase gating, and combined summary. Each phase delegates to its dedicated skill.
For single-dimension checks, use the individual skills directly.
Phase 1: Structure → doc-maintenance
Phase 2: Accuracy → doc-claim-validator
Phase 3: Completeness → doc-completeness-audit
Phase 4: Quality → doc-quality-review
Phase 5: Architecture → doc-architecture-review
Each phase depends on the prior phase being reasonably healthy:
| Phase | Depends on | Reason | |-------|-----------|--------| | Structure | — | Foundation. Broken links and orphans must be visible before anything else | | Accuracy | Structure | Can't verify claims in docs you can't find or that have broken references | | Completeness | Accuracy | No point inventorying gaps if existing docs contain false claims | | Quality | Completeness | Review prose quality only on docs you intend to keep | | Architecture | Completeness | Need to know what's missing before evaluating whether the structure can hold it |
Before running phases, establish what to audit:
docs/, site/, README.md)git diff --name-only main...HEAD -- '*.md')Pass the scope to each phase so they operate on the same set.
Run each phase sequentially. After each phase, evaluate the gate condition.
doc-maintenance)Run: Invoke doc-maintenance on the scoped doc set.
Gate condition: Check the findings.
doc-claim-validator)Run: Invoke doc-claim-validator on the scoped doc set.
Gate condition: Check the findings.
doc-completeness-audit)Run: Invoke doc-completeness-audit on the scoped doc set.
Gate condition: No gate — always continue to quality. Completeness gaps don't affect quality assessment of existing docs.
doc-quality-review)Run: Invoke doc-quality-review on the scoped doc set (excluding any docs
flagged for removal in earlier phases).
Gate condition: No gate — always continue to architecture.
doc-architecture-review)Run: Invoke doc-architecture-review on the scoped doc set. Pass the
completeness gap list from Phase 3 so the architecture review can assess whether
the current structure can accommodate the missing docs.
Gate condition: Final phase — no gate.
After all phases complete (or after an early stop), merge results into a single dashboard.
# Documentation Health Audit
**Audit date:** YYYY-MM-DD
**Scope:** [what was audited]
**Phases completed:** N / 5
**Early stop:** [yes/no — if yes, which phase and why]
---
## Dashboard
| Phase | Skill | Status | Key Metric | Top Finding |
|-------|-------|--------|------------|-------------|
| 1. Structure | doc-maintenance | PASS/WARN/FAIL | N broken links, N orphans | [one line] |
| 2. Accuracy | doc-claim-validator | PASS/WARN/FAIL | N P0, N P1 claims failed | [one line] |
| 3. Completeness | doc-completeness-audit | PASS/WARN/FAIL | N% coverage, N P0 gaps | [one line] |
| 4. Quality | doc-quality-review | PASS/WARN/FAIL | Grade: [A-F], avg score N/25 | [one line] |
| 5. Architecture | doc-architecture-review | PASS/WARN/FAIL | Grade: [A-F], N/35 score | [one line] |
**Overall health:** [Healthy / Needs Attention / Critical]
---
## Priority Actions
Top 5 actions across all phases, ordered by impact:
1. [Phase N] [specific action] — [why it matters]
2. [Phase N] [specific action] — [why it matters]
3. [Phase N] [specific action] — [why it matters]
4. [Phase N] [specific action] — [why it matters]
5. [Phase N] [specific action] — [why it matters]
---
## Phase Reports
[Include or link to each phase's full report below]
### Phase 1: Structure
[doc-maintenance report]
### Phase 2: Accuracy
[doc-claim-validator report]
### Phase 3: Completeness
[doc-completeness-audit report]
### Phase 4: Quality
[doc-quality-review report]
### Phase 5: Architecture
[doc-architecture-review report]
Not every audit needs all five phases. Common partial runs:
| Scenario | Phases to run | |----------|--------------| | "Are the docs accurate?" | 1 + 2 only | | "What's missing?" | 1 + 3 only | | "Is the writing good?" | 4 only | | "Can users find things?" | 1 + 5 only | | "Quick pre-release check" | 1 + 2 + 3 (skip quality and architecture) | | "Full audit" | All 5 |
When running a subset, skip gates for omitted phases.
docs/archive/) — they are historicaldevelopment
Product vision, roadmap development, and go-to-market execution with structured prioritization frameworks. Use when evaluating features, planning product direction, or assessing market fit.
development
Complete operational workflow for implementer agents (Codex, Gemini, etc.) making code changes and writing tests. Drives all work through atomic commits — each loop operates on the smallest complete, reviewable change. Defines the Code Change Loop, Test Writing Loop, Lint Gate, and Issue Filing process with circuit breakers, severity levels, and escalation rules. Requires `cortex git commit` for all commits. Includes bundled provider-aware review scripts that keep same-model shell-outs as the last resort, plus a fresh-context Codex fallback for code review and test audit. Use this skill when starting any implementation task.
development
Use this skill when writing product requirements documents, prioritizing features, creating user stories, defining acceptance criteria, or setting product metrics. Trigger phrases: 'write a PRD for', 'prioritize this feature backlog', 'write user stories for', 'help me define acceptance criteria', 'what metrics should we track for'. Not for writing code, designing UI mockups, or conducting user research interviews.
tools
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.