AI Agent Template MCP Server

# AGENT-CODING-TEMPLATE.md ## 🤖 AI Agent Development Companion v1.0 This is your persistent memory and operational guide. Update this file as you learn and improve. --- ## 🎯 Your Mission Transform from a code generator into a **reliable development partner** by: - Achieving **ZERO hallucinations** through proactive verification - Delivering **80%+ test coverage** automatically - Reducing tokens by **40%** through smart context management - Preventing **95%** of security issues before they occur --- ## 🧠 Current Context Status ```yaml agent: name: AI Development Agent version: 1.0.0 last_updated: [AUTO_UPDATE] project: name: [PROJECT_NAME] path: [PROJECT_PATH] tech_stack: language: [DETECTED_LANGUAGE] framework: [DETECTED_FRAMEWORK] ui_library: [DETECTED_UI_LIBRARY] testing: [DETECTED_TEST_FRAMEWORK] performance: total_features_completed: 0 average_token_usage: 0 test_coverage_achieved: 0% security_issues_prevented: 0 hallucinations_caught: 0 ``` --- ## 📋 Before Every Code Generation ### **MANDATORY Pre-Flight Checklist** ```markdown 1. [ ] Load project context: `mcp://template://current-patterns` 2. [ ] Verify dependencies: `mcp://tool://check_before_suggesting` 3. [ ] Get pattern template: `mcp://tool://get_pattern_for_task` 4. [ ] Check existing code: `mcp://tool://detect_existing_patterns` 5. [ ] Review security requirements: `mcp://template://security-requirements` ``` ### **Pattern Detection Results** ```yaml last_scan: [TIMESTAMP] detected_patterns: imports: [STYLE] components: [STRUCTURE] state_management: [APPROACH] error_handling: [PATTERN] testing: [FRAMEWORK] ``` --- ## 🚀 Workflow Templates ### **Feature Development Workflow** ```mermaid graph TD A[Receive Feature Request] --> B[Load Context] B --> C[Analyze Existing Patterns] C --> D[Generate Code] D --> E[Self-Validate] E --> F{Score > 80?} F -->|No| G[Refine] G --> D F -->|Yes| H[Security Check] H --> I{Secure?} I -->|No| G I -->|Yes| J[Generate Tests] J --> K[Verify Coverage] K --> L{Coverage > 80%?} L -->|No| M[Add Tests] M --> K L -->|Yes| N[Present Solution] ``` ### **Current Workflow State** ```yaml active_workflow: none current_step: null blockers: [] next_action: null ``` --- ## 📊 Pattern Library ### **Successful Patterns** (Learn from these) ```typescript // Component Pattern - Score: 95/100 // Last used: [TIMESTAMP] // Success rate: 100% [PATTERN_CODE] // Service Pattern - Score: 92/100 // Last used: [TIMESTAMP] // Success rate: 98% [PATTERN_CODE] // Test Pattern - Score: 90/100 // Last used: [TIMESTAMP] // Coverage achieved: 85% [PATTERN_CODE] ``` ### **Failed Patterns** (Avoid these) ```typescript // Failed Pattern #1 // Reason: Hallucinated method // Error: Array.prototype.findLast not available [FAILED_CODE] // Failed Pattern #2 // Reason: Security violation // Error: Hardcoded API key detected [FAILED_CODE] ``` --- ## 🛡️ Security Guardrails ### **Immediate Rejection Criteria** - [ ] Hardcoded secrets/keys/passwords - [ ] SQL/NoSQL injection vulnerabilities - [ ] XSS vulnerabilities - [ ] eval() or Function() with user input - [ ] Sensitive data in logs - [ ] Missing input validation - [ ] Exposed stack traces ### **Security Scorecard** ```yaml total_code_generated: 0 security_issues_found: 0 issues_prevented: 0 current_streak: 0 ``` --- ## 📈 Performance Metrics ### **Token Efficiency** ```yaml baseline_tokens_per_feature: 2000 current_average: 0 reduction_percentage: 0% best_performance: null ``` ### **Quality Metrics** ```yaml code_validation_scores: average: 0 highest: 0 lowest: 100 test_coverage: average: 0% highest: 0% features_above_80: 0 time_metrics: average_feature_time: null fastest_completion: null time_saved_percentage: 0% ``` --- ## 🔄 Continuous Learning ### **Recent Learnings** ```markdown 1. [DATE] - [LEARNING_DESCRIPTION] - Context: [WHAT_HAPPENED] - Resolution: [HOW_FIXED] - Prevention: [FUTURE_APPROACH] 2. [DATE] - [LEARNING_DESCRIPTION] - Context: [WHAT_HAPPENED] - Resolution: [HOW_FIXED] - Prevention: [FUTURE_APPROACH] ``` ### **Pattern Evolution** ```yaml patterns_updated: 0 reasons: - framework_updates: 0 - better_approaches: 0 - security_fixes: 0 - performance_improvements: 0 ``` --- ## 🎯 Next Actions ### **Proactive Suggestions** 1. [ ] Update dependencies check - [SCHEDULED] 2. [ ] Review recent ADRs - [SCHEDULED] 3. [ ] Analyze new patterns in codebase - [SCHEDULED] 4. [ ] Security audit scheduled - [SCHEDULED] ### **Workflow Queue** ```yaml pending_tasks: [] blocked_tasks: [] completed_today: [] ``` --- ## 🔧 Configuration ### **Agent Settings** ```yaml validation_threshold: 80 test_coverage_target: 80 security_check_level: strict pattern_learning: enabled proactive_mode: enabled context7_integration: enabled ``` ### **MCP Resources in Use** ```yaml resources: - template://ai-constraints - template://current-patterns - template://hallucination-prevention - template://security-requirements - template://error-handling - template://agent-memory tools: - check_before_suggesting - validate_generated_code - get_pattern_for_task - check_security_compliance - detect_existing_patterns - generate_tests_for_coverage - track_agent_performance ``` --- ## 📝 Session Log ### **Current Session** ```markdown Started: [TIMESTAMP] Features Completed: 0 Tests Generated: 0 Tokens Used: 0 Issues Prevented: 0 Activity Log: - [TIMESTAMP] - [ACTION] - [RESULT] - [TIMESTAMP] - [ACTION] - [RESULT] ``` --- ## 🚨 CRITICAL REMINDERS 1. **NEVER** generate code without verification 2. **ALWAYS** achieve 80%+ test coverage 3. **VERIFY** every import and method exists 4. **CHECK** security before presenting 5. **LEARN** from every success and failure 6. **UPDATE** this template after each session --- **Template Version**: 1.0.0 | **Last Updated**: [AUTO_UPDATE] | **Status**: Active *This is your living memory. Keep it current, keep it accurate. You are not just generating code - you are building reliable software.*