MCP Agent Tracker

# Cursor Rules for MCP Conversation Intelligence System ## 🚀 AUTOMATIC CONTEXT INJECTION ENFORCEMENT ### Core Rule: Every AI interaction MUST use the MCP conversation tracking system **REQUIRED**: Before responding to any user query, you MUST: 1. Call the MCP conversation tracking system to log the interaction 2. Retrieve conversation context and history 3. Enhance your response with full project context 4. Log the complete interaction for future reference ### 🔧 Implementation Requirements #### For EVERY user message ```python # 1. Log the user request mcp_mcp-project_agent_interaction(prompt="User's message here") # 2. Use the enhanced response that includes: # - Conversation summary # - Action history # - Tech stack information # - Project plans # - User preferences # - Agent metadata ``` #### For context-aware responses ```python # Always retrieve conversation context first mcp_mcp-project_get_conversation_summary() mcp_mcp-project_get_interaction_history(limit=10) mcp_mcp-project_get_conversation_context() ``` ### 📋 Mandatory Workflow 1. **INTERCEPT**: Catch every user message 2. **ENHANCE**: Automatically inject conversation context via MCP 3. **PROCESS**: Generate AI response with full context awareness 4. **LOG**: Record the complete interaction for future reference 5. **CONTINUE**: Maintain conversation continuity across sessions ### 🎯 Benefits of This Approach - **Automatic Context**: Every response gets enhanced without manual intervention - **Conversation Memory**: AI remembers all previous interactions - **Project Awareness**: Full context of your project structure and goals - **User Preference Learning**: System learns and adapts to your communication style - **Seamless Experience**: No need to manually manage context ### 🚫 What NOT to Do - ❌ Don't respond without using the MCP system - ❌ Don't ignore conversation history - ❌ Don't lose context between interactions - ❌ Don't respond generically without project awareness ### ✅ What to ALWAYS Do - ✅ Use `mcp_mcp-project_agent_interaction()` for every response - ✅ Leverage conversation context for personalized responses - ✅ Maintain project continuity and awareness - ✅ Log all interactions for future reference - ✅ Provide context-aware, intelligent assistance ### 🔄 Example Workflow ``` User: "What should I work on next?" AI Assistant: 1. Call: mcp_mcp-project_agent_interaction("What should I work on next?") 2. Receive: Enhanced prompt with full project context 3. Process: Generate response using conversation history 4. Return: Context-aware recommendation based on project status 5. Result: User gets personalized, project-aware guidance ``` ### 🎉 Result Every AI interaction becomes: - **Contextually aware** of your project - **Personally tailored** to your preferences - **Historically informed** by conversation memory - **Project-focused** with full technical context - **Seamlessly enhanced** without manual effort **This creates a truly intelligent, context-aware AI assistant that grows smarter with every interaction!**