Janitor Report Agent Skill

Health Report

Generate a comprehensive health report combining inventory, quality checks, duplicate detection, and broken skill findings.

Overview

One report covering every place a skill lives: user, project, codex, plugin, and source scopes. Plugin skills appear with their full invocation name (e.g. marketing-skills:image, figma:figma-use) — these were invisible to the v1.2 report. The scan JSON also includes an agents array (subagents from ~/.claude/agents — their descriptions are always-loaded too) and per-plugin update_available (installed commit vs marketplace HEAD) — surface stale plugins and heavy agents in the report.

Modes:

• /janitor-report (default) — full health check: inventory + lint + duplicates + broken
• /janitor-report --brief — inventory only (replaces the old /janitor-audit)

Prerequisites

• Claude Code with the skills-janitor plugin installed (provides scripts/scan.sh, lint.sh, detect_dupes.sh)
• bash 3.2+ (the stock macOS bash works; no external dependencies)
• Read access to ~/.claude/skills, ~/.claude/agents, and installed plugin metadata
• No authentication or API keys required — the scripts read local files only

Instructions

Step 1: Run the scripts

Full report:

bash ~/.claude/skills/skills-janitor/scripts/scan.sh
bash ~/.claude/skills/skills-janitor/scripts/lint.sh
bash ~/.claude/skills/skills-janitor/scripts/detect_dupes.sh

Brief mode (inventory only):

bash ~/.claude/skills/skills-janitor/scripts/scan.sh

Step 2: What each script covers

• Inventory (scan.sh) — all skills across scopes, symlink status, frontmatter fields, line counts; agents and plugins (with update_available) arrays; broken_symlinks count.
• Quality checks (lint.sh) — Critical: broken symlinks, missing SKILL.md, missing frontmatter. Warning: missing/empty name or description, description too short/long, missing version. Info: no body content, no Gotchas section, large files.
• Duplicate detection (detect_dupes.sh) — name collisions (two distinct skills with the same qualified name at different paths) and description overlap (Jaccard similarity >30%), with cross-scope user-vs-plugin pairs explicitly surfaced (e.g. marketing-seo-audit (user) ↔ marketing-skills:seo-audit (plugin)).

Step 3: Present a unified report

Merge the three outputs into one table with severity levels, then recommend actions.

Output

| Skill                            | Scope    | Status      | Issues                                   |
|----------------------------------|----------|-------------|------------------------------------------|
| marketing-copywriting            | user     | DUPLICATE?  | 90% overlap with marketing-skills:copywriting |
| seo-audit                        | user     | WARNING     | Description too short                    |
| old-deploy-helper                | user     | CRITICAL    | Broken symlink                           |
| figma:figma-use                  | plugin   | OK          | -                                        |

Recommended actions per issue type:

• Broken symlinks → /janitor-fix --prune
• Quality issues → /janitor-fix
• User-vs-plugin duplicates → uninstall the user-scope copy and rely on the plugin (or vice versa)
• Stale plugins (update_available: true) → /plugin update <name>
• Token waste → /janitor-value
• Suspicious content (injection phrases, dangerous scripts) → /janitor-security
• Unused MCP servers (run scripts/mcp.sh for the inventory) → remove the config entry or /janitor-swipe

Error Handling

• Error: scan.sh: No such file or directory

Solution: The plugin is installed under a different root. Locate it with ls ~/.claude/skills or check the plugin cache, then run the scripts from their actual location.

• Error: Scan output is not valid JSON

Solution: Re-run with stderr visible (drop any 2>/dev/null) and report the failing line; a skill directory with unusual characters is the usual cause.

• Error: plugins array is empty despite installed plugins

Solution: ~/.claude/plugins/installed_plugins.json is missing or in an unknown format — report which Claude Code version is in use.

Examples

Example 1: Full health check

Input: "Check my skills for problems."

Output: Run all three scripts, merge into the severity table, and lead with counts: "175 skills scanned: 2 critical (broken symlinks), 5 warnings, 3 likely duplicates, 2 plugins have updates available."

Example 2: Inventory only

Input: "/janitor-report --brief"

Output: Run scan.sh only and present the inventory grouped by scope with per-scope counts.

Resources

• Scripts (plugin-relative): {baseDir}/../../scripts/scan.sh, lint.sh, detect_dupes.sh
• /janitor-fix — auto-fix the issues this report finds
• /janitor-value — token cost + usage (combined)
• /janitor-discover — find or evaluate new skills