Unified Docs Hub

🚀 Unified Docs Hub - The Ultimate MCP Documentation Server

Transform your AI assistant into a documentation powerhouse! Unified Docs Hub is an MCP (Model Context Protocol) server that creates a massive, searchable knowledge base from 170+ curated repositories and 1000+ auto-discovered GitHub projects.

Ever wished your AI assistant had instant access to ALL the documentation it needs? This MCP server solves that by:

- 📚 Massive Knowledge Base: 170+ hand-picked repositories + 1000+ auto-discovered popular projects
- 🔍 Lightning-Fast Search: Full-text search across 11,000+ documentation files in milliseconds
- 🤖 AI-Optimized: Perfect for Claude, ChatGPT, and other AI assistants using MCP
- 📈 Self-Updating: Automated daily updates and weekly discovery of new repositories
- 🎯 Specialized Coverage: Deep expertise in Trading/Finance, AI/ML, DevOps, and 20+ categories

AI: "Show me how to build a crypto trading bot with backtesting" You: unified_search(query="crypto trading bot backtesting", category="Trading & Finance") Result: Instant access to documentation from: - Freqtrade (advanced crypto trading bot) - Backtrader (backtesting framework) - CCXT (100+ exchange APIs) - TA-Lib (200+ technical indicators)

AI: "Explain Kubernetes deployment strategies" You: unified_search(query="kubernetes deployment strategies", category="Cloud/DevOps") Result: Documentation from: - Official Kubernetes docs - Helm charts best practices - ArgoCD GitOps workflows - Istio service mesh patterns

AI: "Set up an MLOps pipeline with experiment tracking" You: unified_search(query="mlops pipeline experiment tracking", category="MLOps") Result: Comprehensive guides from: - MLflow (experiment tracking) - Kubeflow (distributed training) - DVC (data versioning) - Weights & Biases (visualization)

- 🔥 Full-Text Search: SQLite FTS5 engine for sub-second searches across millions of lines
- 📈 Quality Scoring: Curated repos ranked by documentation quality (1-10 scale)
- 🏷️ Smart Categorization: Browse by technology area or programming language
- 🔄 Auto-Discovery: Continuously finds new popular repositories (10k+ stars)
- 💾 Efficient Storage: Deduplication and compression keep the database lean
- 🛡️ Rate Limit Handling: Respects GitHub API limits with smart throttling

- Python 3.8 or higher
- GitHub Personal Access Token (optional but recommended)
- An MCP-compatible AI assistant (Claude Desktop, Continue.dev, etc.)

git clone https://github.com/yourusername/unified-docs-hub.git cd unified-docs-hub

python3 -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate pip install -r requirements.txt

For Claude Desktop, add to~/Library/Application Support/Claude/claude_desktop_config.json:

{ "mcpServers": { "unified-docs-hub": { "command": "/path/to/unified-docs-hub/venv/bin/python", "args": ["/path/to/unified-docs-hub/unified_docs_hub_server.py"], "env": { "GITHUB_TOKEN": "your-github-token-here" } } } }

- Initial indexing(optional - the server will do this automatically)

# Index all curated repositories python -c "import asyncio; from unified_docs_hub_server import index_repositories; asyncio.run(index_repositories('smart'))"

Search across all documentation with powerful filters.

# Basic search unified_search("react hooks tutorial") # Advanced search with filters unified_search( query="transformer architecture attention", category="AI/ML", min_stars=5000 ) # Trading-specific search unified_search( query="options greeks volatility smile", category="Trading & Finance" )

Control repository indexing and discovery.

# Smart mode: Index curated + discover popular (recommended) index_repositories(mode="smart") # Update all existing repos index_repositories(mode="update") # Discover new trending repos index_repositories(mode="discover", min_stars=5000, count=50)

# List all Trading & Finance repos list_repositories(category="Trading & Finance") # Show only curated high-quality repos list_repositories(source="curated", limit=20)

Get all documentation for a specific repository.

# Get all Kubernetes docs get_repository_docs("kubernetes/kubernetes") # Get trading library docs get_repository_docs("freqtrade/freqtrade")

get_statistics() # Returns: Total repos, documents, categories, languages, API status

The server includes automated indexing that keeps your knowledge base fresh:

# Run the setup script ./setup_automated_indexing.sh # Or manually start the updater python automated_index_updater.py --once # Run once python automated_index_updater.py # Run continuously

- Daily: Updates all curated repositories (2 AM, 2 PM)
- Weekly: Discovers new trending repositories
- On-Demand: Manual updates via MCP tools

unified-docs-hub/ ├── unified_docs_hub_server.py # Main MCP server ├── database.py # SQLite + FTS5 engine ├── github_client.py # GitHub API integration ├── response_limiter.py # HTTP/2 error prevention ├── repositories.yaml # Curated repo list ├── automated_index_updater.py # Auto-update system └── unified_docs.db # Documentation database

- Curation: Hand-picked repositories inrepositories.yamlwith quality scores
- Discovery: Automatically finds popular repos (10k+ stars) via GitHub API
- Indexing: Downloads and indexes README, docs/, and documentation files
- Storage: SQLite with FTS5 for efficient full-text search
- Serving: FastMCP server provides tools for AI assistants
- Updates: Automated system keeps documentation current

- Instant access to ML framework documentation
- Compare different approaches across libraries
- Find code examples and best practices

- Complete algorithmic trading documentation
- Options pricing models and strategies
- Backtesting frameworks and market data APIs

- Kubernetes patterns and anti-patterns
- CI/CD pipeline examples
- Infrastructure as Code templates

- Frontend framework comparisons
- Backend architecture patterns
- Database optimization techniques

curated_repositories: - repo: "owner/awesome-project" category: "Web Development" description: "An awesome web framework" quality_score: 9 priority: high doc_paths: - "docs/" - "README.md" topics: ["web", "framework", "javascript"]

Add new categories to group related technologies:

- repo: "quantum-computing/qiskit" category: "Quantum Computing" # New category! description: "Quantum computing SDK"

See our journey of building this massive knowledge base:

- EXPANSION_SUMMARY.md- Overview of all expansions
- TRADING_KNOWLEDGE_BASE_COMPLETE.md- Trading & Finance deep dive
- ULTIMATE_TRADING_EXPANSION.md- Final trading expansion details
- FINAL_EXPANSION_REPORT_2025.md- Complete 2025 expansion

We welcome contributions! Please see ourContributing Guidefor details.

This project is licensed under the MIT License - see theLICENSEfile for details.

- Model Context Protocolfor enabling AI-assistant integrations
- All the amazing open-source projects indexed in our knowledge base
- The GitHub API for making documentation discovery possible

GitHits MCP gives AI agents access to open-source code search, package docs, real-world examples, dependency metadata, changelogs, and vulnerability context for better software development decisions.