MCP Prompt Optimizer

# Custom Slash Commands This directory contains custom Claude Code slash commands for the MCP Prompt Optimizer project. ## Available Commands ### `/ori` - Optimize Research Implement **Purpose:** Autonomous workflow for researching, validating, and implementing changes with minimal user input. **Usage:** ```bash /ori <your request> ``` **Examples:** ```bash # Add a new feature /ori add JWT authentication to the Express API # Fix a bug /ori fix the memory leak in the data processing pipeline # Optimize performance /ori optimize database queries for the user dashboard # Research and implement best practices /ori add rate limiting to all API endpoints using industry standards ``` **What it does:** 1. **Phase 1: Research** - Automatically searches web, docs, and codebase 2. **Phase 2: Verify** - Cross-validates findings and checks for risks 3. **Phase 3: Implement** - Applies changes with error handling 4. **Phase 4: Document** - Updates README, CHANGELOG, and other docs **Configuration:** Edit `.claude/ori-config.json` to customize behavior: - Auto-approve low-risk changes - Research depth and source count - Documentation update preferences - Error handling settings **Full Documentation:** See [ori.md](ori.md) --- ## Creating Custom Commands To create a new slash command: 1. Create a new `.md` file in this directory: `.claude/commands/your-command.md` 2. Write the command specification/prompt 3. Use the command: `/your-command <args>` **Example:** ```bash # Create .claude/commands/test.md with content: # "Run all tests in the project and report results" # Then use it: /test ``` --- ## Command Best Practices 1. **Be specific** - Clear commands get better results 2. **Provide context** - Include tech stack, constraints when relevant 3. **Review output** - Check implementation before committing 4. **Use config** - Customize behavior via config files 5. **Iterate** - Refine commands based on results --- **Last updated:** 2025-11-08