Model Context Protocol Server

# Environment Variables Reference Create a `.env` file in the project root with these environment variables to configure the application. ## API Keys and External Services ``` # Exa API for web search (required for search functionality) EXA_API_KEY=your_exa_api_key_here # FireCrawl API for web content crawling (required for RAG functionality) FIRECRAWL_API_KEY=your_firecrawl_api_key_here # OpenAI API for embeddings and LLM (alternative to Ollama) OPENAI_API_KEY=your_openai_api_key_here OPENAI_API_BASE=https://api.openai.com/v1 # Optional: custom endpoint OPENAI_EMBEDDING_MODEL=text-embedding-ada-002 OPENAI_MODEL=gpt-3.5-turbo # or gpt-4, etc. ``` ## Ollama Configuration (Local LLM and Embeddings) ``` # Ollama configuration for local LLM and embeddings OLLAMA_BASE_URL=http://localhost:11434 OLLAMA_MODEL=mistral:latest # or any other model you have downloaded ``` ## MCP Server Configuration ``` # MCP server configuration MCP_HOST=localhost MCP_PORT=8000 ``` ## Logging Configuration ``` # Logging level LOG_LEVEL=INFO # Options: DEBUG, INFO, WARNING, ERROR, CRITICAL ``` ## Example .env file Here's an example of a complete `.env` file: ``` # API Keys EXA_API_KEY=exa-abc123def456 FIRECRAWL_API_KEY=fca-xyz789abc123 # OpenAI OPENAI_API_KEY=sk-abc123def456 OPENAI_EMBEDDING_MODEL=text-embedding-ada-002 OPENAI_MODEL=gpt-3.5-turbo # Ollama OLLAMA_BASE_URL=http://localhost:11434 OLLAMA_MODEL=mistral:latest # MCP Server MCP_HOST=localhost MCP_PORT=8000 # Logging LOG_LEVEL=INFO ``` ## Using the Environment Variables The application will automatically load these variables when you run it. You can also specify a custom `.env` file path with the `--env-file` command line option.