docslisted
Install: claude install-skill Smart-AI-Memory/attune-ai
# docs
Documentation generation, explanation, and accuracy auditing.
## Routes
| Subcommand | Action |
| ---------- | ------ |
| `audit` | Quick inline accuracy checks (10 checks) |
| `audit-deep` | Full DocAuditWorkflow with auto-fix |
| `explain` | Explain how code works |
| `generate` | Generate documentation |
| `readme` | Update README |
| `changelog` | Generate changelog |
| `overview` | High-level project overview |
## Usage
```bash
/docs # Ask what to do
/docs audit # Verify doc accuracy
/docs explain # Explain code
/docs generate # Generate documentation
/docs readme # Update README
/docs changelog # Generate changelog
```
## Behavior
### audit
Cross-reference documentation claims against the actual
codebase. Run all checks below and report a summary table
with pass/fail per check and file:line references for any
mismatches.
#### 1. Test count
Run:
```bash
pytest --collect-only -q 2>/dev/null | tail -1
```
Compare the collected/passing count against:
- The static badge in README.md (the number in the
`img.shields.io/badge/tests-` URL)
- Any "X+ tests" claims in README body text
- Any test counts in CHANGELOG.md (current release only)
Flag if the README count exceeds the actual count by
more than 5%.
#### 2. Workflow count
Count entries under `[project.entry-points."attune.workflows"]`
in `pyproject.toml`. Compare against any "N built-in" claims
in README.md. Flag mismatches.
#### 3