# Commands This directory contains all slash commands for the compounding-engineering plugin. Each `.md` file defines a command that can be invoked via `claude /command-name`. ## Documentation Management ### `/release-docs` **Purpose:** Build and update the documentation site with current plugin components. **Usage:** ```bash # Full documentation release claude /release-docs # Preview changes without writing files claude /release-docs --dry-run ``` **What it does:** 1. Inventories all current components (agents, commands, skills, MCP servers) 2. Updates `docs/index.html` with accurate stats 3. Regenerates reference pages (`agents.html`, `commands.html`, `skills.html`, `mcp-servers.html`) 4. Updates `changelog.html` with latest version history 5. Ensures counts in `plugin.json` and `marketplace.json` match actual files 6. Validates all JSON files **When to run:** - After adding, removing, or modifying any agent - After adding, removing, or modifying any command - After adding, removing, or modifying any skill - After adding, removing, or modifying any MCP server - Before releasing a new version ## Workflow Commands ### `/plan_review` Multi-agent plan review running in parallel for thorough analysis. ### `/resolve_parallel` Resolve TODO comments in the codebase in parallel. ### `/resolve_pr_parallel` Resolve PR comments in parallel. ### `/resolve_todo_parallel` Resolve TODO items from a list in parallel. ## Development Commands ### `/changelog` Create engaging changelogs for recent merges to main branch. ### `/generate_command` Generate new slash command files from a description. ### `/create-agent-skill` Create or edit Claude Code skills with best practices. ### `/heal-skill` Fix skill documentation issues and formatting. ### `/prime` Prime/setup command for initializing projects. ### `/reproduce-bug` Reproduce bugs using logs and console output. ### `/report-bug` Report bugs in the compounding-engineering plugin with structured workflow. ### `/triage` Triage and prioritize issues. ## Command File Structure Each command file follows this structure: ```markdown --- name: command-name description: Brief description of what the command does argument-hint: "[optional arguments description]" --- # Command Title Instructions for Claude on how to execute this command... ``` ## Adding a New Command 1. Create a new `.md` file in this directory 2. Add the frontmatter with `name`, `description`, and optional `argument-hint` 3. Write detailed instructions for Claude 4. Run `/release-docs` to update documentation 5. Test with `claude /your-command-name` ## Best Practices - Keep command names short and descriptive (use hyphens, not underscores) - Provide clear step-by-step instructions - Include examples of expected output - Document any prerequisites or dependencies - Use parallel agent invocation when tasks are independent