v4.5.0 - API-First MCP Architecture

🚀 API-First MCP Architecture & Legacy Cleanup

New Features

API-First MCP Server (mcp_server/server_api_client.py)

• Thin HTTP client calls FastAPI backend (468 lines)
• 5 MCP tools via HTTP API:
  • check_status - API health and connectivity
  • search_standards - Search with filters
  • analyze_code - Code analysis
  • list_standards - List all standards
  • get_recommendations - Improvement suggestions
• Clean stdout (MCP protocol compliant - no Neo4j pollution)
• Multi-client support (Claude Desktop, Claude Code, other agents)
• Remote access capability via HTTP

Architecture Improvements

Before: MCP → Direct Neo4j/Files → Stdout pollution, single-client
After: MCP Client → HTTP API → FastAPI → Neo4j → Clean stdout, multi-client

Benefits:

• Remote API access (not just local file-based)
• Centralized authentication and rate limiting
• Redis caching for improved performance
• 3,420 standards accessible via Neo4j
• Graceful degradation when services unavailable

Documentation

• API_FIRST_MCP_IMPLEMENTATION.md (321 lines): Architecture, testing, next steps
• MCP_SERVER_ARCHITECTURE_ANALYSIS.md (303 lines): Server evolution, Issue #11 investigation
• .mcp.json: Claude Desktop configuration

Cleanup

• Archived 8 legacy MCP server files to mcp_server/archive/
• Organized test files into tests/integration/
• Resolves GitHub Issue #11 (MCP server confusion)

Two MCP Server Options

1. server_simple.py (Local, file-based) - For simple deployments
2. server_api_client.py (Remote, HTTP API) - NEW - For remote access

Both are production-ready and MCP protocol compliant.

Testing

• API server verified running on port 8000
• Health endpoint healthy
• Neo4j connected (3,420 standards)
• Redis connected (caching available)
• HTTP endpoints returning 200 OK

Backward Compatibility

• No breaking changes
• server_simple.py still available for local use
• All existing functionality preserved

---

See DEVELOPMENT_STATE.md for detailed implementation notes.

v4.2.2 - Auto-Refresh Standards Feature

Release v4.2.2 - Auto-Refresh Standards Feature

🎉 Major Feature Release

This release implements automatic refresh of coding standards that are older than 30 days, ensuring developers always receive current best practices. The feature uses deep research mode with iterative AI refinement to achieve exceptional quality (8.5-9.5/10).

✨ Key Features

Auto-Refresh Standards on Access

• Automatic Staleness Detection: Standards older than 30 days trigger refresh automatically
• Deep Research Mode: Multi-pass AI generation with self-critique for quality
• Quality Threshold: Enforces minimum 8.5/10 quality score
• Background Processing: Non-blocking updates with configurable worker pool
• Comprehensive Metrics: 5 new API endpoints for monitoring and management

Production Testing Results ✅

Successfully validated by updating the 3 oldest standards:

Standard	Quality Score	Iterations	Content Growth
Python	9.0/10	1	+28% (12KB → 16KB)
Java	9.0/10	2	+267% (8KB → 29KB)
General	9.5/10	2	+102% (11KB → 22KB)

Highlights:

• Deep research demonstrated self-improvement (Java & General: 8.0 → 9.0+)
• All standards exceeded quality threshold
• Automatic version bumps to v2.0.0
• November 2025 best practices applied

📦 Implementation Details

New Code (2,475 lines)

Production:

• StandardsAccessService - 605 lines (core access layer)
• MetricsRouter - 310 lines (5 monitoring endpoints)
• Background task queue with retry logic
• 7 new configuration settings

Testing:

• 27 unit tests (100% pass rate)
• 61.26% coverage on new service
• 2 integration test scripts
• Overall coverage: 13.51% → 16.60%

Documentation:

• Complete architecture docs (AUTO_REFRESH_DESIGN.md)
• API documentation for all endpoints
• Configuration guides and examples

⚙️ Configuration

New environment variables (all optional with defaults):

ENABLE_AUTO_REFRESH_ON_ACCESS=true          # Enable/disable
STANDARD_FRESHNESS_THRESHOLD_DAYS=30        # Age threshold
AUTO_REFRESH_MODE=background                # blocking/background
AUTO_REFRESH_MAX_CONCURRENT=3               # Worker pool size
AUTO_REFRESH_RETRY_ATTEMPTS=2               # Retry logic
AUTO_REFRESH_RETRY_DELAY_SECONDS=60         # Retry delay
AUTO_REFRESH_USE_DEEP_RESEARCH=true         # Iterative refinement

🔌 New API Endpoints

GET    /api/v1/metrics/auto-refresh                              # Overall metrics
GET    /api/v1/metrics/standards/{id}/refresh-status            # Per-standard status
PATCH  /api/v1/metrics/standards/{id}/auto-refresh-settings     # Update config
POST   /api/v1/metrics/standards/{id}/refresh                   # Manual trigger
GET    /api/v1/metrics/health                                   # Health check

🚀 Upgrade Instructions

Requirements

• Python 3.11+
• Gemini API key (for deep research mode)

Installation

# Pull latest code
git pull origin main
git checkout v4.2.2

# Install dependencies (if updated)
pip install -r requirements.txt

# Optional: Configure settings in .env
ENABLE_AUTO_REFRESH_ON_ACCESS=true
STANDARD_FRESHNESS_THRESHOLD_DAYS=30

Verification

# Run tests
pytest tests/unit/services/test_standards_access_service.py -v

# Check feature status
curl http://localhost:8000/api/v1/metrics/health

📝 Breaking Changes

None - Fully backward compatible!

🐛 Bug Fixes & Security

• Removed hardcoded GitHub tokens from CLAUDE.md
• Cleaned git history of sensitive data
• Enhanced security documentation

📚 Documentation

• Auto-Refresh Design
• Development State
• Session Summary
• Issue Resolution

🙏 Credits

Closes #8 - Auto-refresh standards older than 30 days on access

---

Full Changelog: v4.2.1...v4.2.2

🤖 Generated with Claude Code

Claude Integrated MCP Services for code standards and auditing

Functional release with MCP integration