Audit Commands Command

Audit Claude Code slash commands for quality, compliance, and maintainability.

Initialization

Before auditing, initialize the environment:

Get the current UTC date, capture the project root path, ensure the temp directory exists, and clean up stale audit files (unless --recover). The command-development skill provides authoritative validation guidance (auto-loaded when this command runs).

What Gets Audited

• YAML frontmatter (description, allowed-tools, argument-hint)
• Description quality (clear, concise, action-oriented)
• Tool restrictions appropriateness
• Argument handling patterns ($ARGUMENTS, $1, $2)
• File naming and organization
• Content structure and quality

Command Arguments

Argument	Description
(none)	Smart mode: audit modified, never-audited, or stale (>90 days) commands
--force	Audit ALL commands regardless of status
--plugin-only	Only audit local plugin commands
--project-only	Only audit project commands (.claude/commands/)
--global-only	Only audit globally installed plugin commands
--recover	Recover results from interrupted audit
command-name	Audit specific command(s) by name

Step 1: Discover Command Sources

Detect all command sources in local repo and globally installed plugins.

For local discovery, check marketplace repos, single plugin repos, and .claude/commands/. Track plugin names for deduplication.

For global discovery, check ~/.claude/plugins/ (Unix) or %USERPROFILE%\.claude\plugins\ (Windows). Skip globals with local dev versions.

Step 2: Parse Arguments

Parse flags and command names. Read audit logs for each source.

Step 3: Present Audit Plan

Display mode, sources discovered, deduplication status, and audit queue with batching strategy.

Step 4: Execute Audits

For each command, spawn the command-auditor subagent with source, path, and last audit date. Run in parallel batches of 3-5.

Use progressive persistence: persist results after each batch. Subagents write to .claude/temp/. The main conversation thread collects results and updates audit logs using its Write/Edit tools.

Step 5: Final Summary

Report total audited by source, results, and details table. Note that global command fixes must be applied manually.

Scoring

Category	Points
File Structure	20
YAML Frontmatter	25
Description Quality	20
Tool Configuration	15
Content Quality	20

Thresholds: 85+ = PASS, 70-84 = WARNINGS, <70 = FAIL

Audit Log Location

All audit results are written to .claude/audit/commands.md.

Use /audit-log commands to view current audit status.

Recovery

If context collapses mid-audit, use /audit-commands --recover to recover from .claude/temp/ files.

Example Usage

Example 1: Audit All Commands

User: /audit-commands

Claude: Discovering command sources...

## Audit Plan
**Mode**: SMART
- Plugin: claude-ecosystem (24 commands)
- Plugin: code-quality (4 commands)
- Deduplicated: claude-ecosystem (global skipped)

**Will audit**: 8 commands in 2 batches

[Spawns command-auditor subagents]

## Audit Complete
| Source | Command | Result | Score |
| --- | --- | --- | --- |
| plugin | scrape-docs | PASS | 100/100 |

Example 2: Audit Specific Command

User: /audit-commands plugin:scrape-docs
Claude: PASS (Score: 100/100)