Claude Code Inventory
Generate a comprehensive inventory of all Claude Code extensions. Execute these steps:
1. Scan for Slash Commands
User commands:
bash
for f in ~/.claude/commands/*.md; do [ -f "$f" ] || continue name=$(basename "$f" .md) desc=$(grep -m1 "^description:" "$f" 2>/dev/null | sed 's/description: *//' || echo "-") echo "USER|$name|$(basename $f)|$desc" done
Project commands:
bash
for f in .claude/commands/*.md; do [ -f "$f" ] || continue name=$(basename "$f" .md) desc=$(grep -m1 "^description:" "$f" 2>/dev/null | sed 's/description: *//' || echo "-") echo "PROJECT|$name|$(basename $f)|$desc" done
2. Scan for Skills
User skills:
bash
for d in ~/.claude/skills/*/; do [ -d "$d" ] || continue name=$(basename "$d") skill_file="$d/SKILL.md" desc=$(grep -m1 "^description:" "$skill_file" 2>/dev/null | sed 's/description: *//' || echo "-") echo "USER|$name|SKILL.md|$desc" done
Project skills:
bash
for d in .claude/skills/*/; do [ -d "$d" ] || continue name=$(basename "$d") skill_file="$d/SKILL.md" desc=$(grep -m1 "^description:" "$skill_file" 2>/dev/null | sed 's/description: *//' || echo "-") echo "PROJECT|$name|SKILL.md|$desc" done
3. Scan for Custom Subagents
User subagents:
bash
for f in ~/.claude/subagents/*.md; do [ -f "$f" ] || continue name=$(basename "$f" .md) desc=$(grep -m1 "^description:" "$f" 2>/dev/null | sed 's/description: *//' || echo "-") echo "USER|$name|$(basename $f)|$desc" done
Project subagents:
bash
for f in .claude/subagents/*.md; do [ -f "$f" ] || continue name=$(basename "$f" .md) desc=$(grep -m1 "^description:" "$f" 2>/dev/null | sed 's/description: *//' || echo "-") echo "PROJECT|$name|$(basename $f)|$desc" done
4. Scan for Hooks
User hooks: Read ~/.claude/settings.json and extract any hooks configuration.
Project hooks: Read .claude/settings.json and .claude/settings.local.json and extract any hooks configuration.
5. Check Installed Plugins
Read ~/.claude/plugins/installed_plugins.json and list enabled plugins with their commands.
Also check ~/.claude/settings.json for enabledPlugins to show status.
6. Output Format
Present the results as a clean inventory table. Use this exact format:
code
┌─────────────────────────────────────────────────────────────────────────────────────────────────┐ │ CLAUDE CODE INVENTORY │ └─────────────────────────────────────────────────────────────────────────────────────────────────┘ SLASH COMMANDS ────────────────────────────────────────────────────────────────────────────────────────────────── Name Scope Source Description ────────────────────────────────────────────────────────────────────────────────────────────────── /command-name User command-name.md Description text here /other-command Project other-command.md Another description SKILLS ────────────────────────────────────────────────────────────────────────────────────────────────── Name Scope Source Description ────────────────────────────────────────────────────────────────────────────────────────────────── skill-name User SKILL.md Skill description SUBAGENTS ────────────────────────────────────────────────────────────────────────────────────────────────── Name Scope Source Description ────────────────────────────────────────────────────────────────────────────────────────────────── Explore Built-in - Fast read-only codebase exploration Plan Built-in - Research and planning in plan mode general-purpose Built-in - Complex multi-step tasks with exploration claude-code-guide Built-in - Claude Code documentation lookup frontend-expert Built-in - Frontend development specialist backend-architect Built-in - Backend and API design specialist fullstack Built-in - Full-stack TypeScript/React implementation supabase Built-in - Supabase backend services specialist debugger Built-in - Error investigation and debugging architect Built-in - System design for AI-native solutions code-reviewer Built-in - Code quality and security review qa Built-in - Quality assurance and DevOps unittest Built-in - Unit test generation projm Built-in - Project management and coordination llmbrains Built-in - LLM engineering and prompt design pr-tester Built-in - Pull request testing statusline-setup Built-in - Status line configuration custom-agent User custom.md Custom agent description HOOKS ────────────────────────────────────────────────────────────────────────────────────────────────── Name Scope Source Description ────────────────────────────────────────────────────────────────────────────────────────────────── pre-commit Project settings.json Runs before git commits PLUGINS (ENABLED) ────────────────────────────────────────────────────────────────────────────────────────────────── Name Commands ────────────────────────────────────────────────────────────────────────────────────────────────── nextjs-vercel-pro@claude-code-templates /nextjs-vercel-pro:nextjs-scaffold, ... performance-optimizer@claude-code-templates /performance-optimizer:performance-audit documentation-generator@claude-code-templates /documentation-generator:update-docs BUILT-IN KEYWORDS ────────────────────────────────────────────────────────────────────────────────────────────────── Name Description ────────────────────────────────────────────────────────────────────────────────────────────────── think / think hard / think harder / ultrathink Extended thinking (increasing depth)
Important Notes
- •Sort by: Type (grouped) > Scope (Built-in first, then User, then Project) > Alphabetical
- •Truncate descriptions longer than 50 characters with "..."
- •Show "-" for missing descriptions
- •Only show sections that have items (skip empty sections)
- •Built-in subagents - Always include the standard set (Explore, Plan, general-purpose, etc.)
- •Built-in keywords - Always show as the final section
Execute the scans silently, then output ONLY the formatted table. No explanatory text before or after.