This is not an MCP server, it's a collection of LangChain/Anthropic SDK scripts. No MCP protocol implementation detected. Tools are defined in two incompatible ways (LangChain @tool decorators in tool_uses.py; raw Anthropic tool dicts in tools.py) with severe duplication: get_weather appears 4 times with different signatures, homelab_status appears 3 times. No tool registration, no unified schema, no error handling. Descriptions are minimal (10-50 chars). Parameters lack type constraints and validation. No output schemas documented. This cannot function as an MCP server.

• Implement MCP server using official MCP SDK (Python: mcp package). Replace LangChain/Anthropic SDK direct calls with MCP transport layer (HTTP+SSE or Streamable HTTP).
• Consolidate duplicate tools: merge 4 get_weather variants into one tool with optional parameters (city: string | null, latitude: number | null, longitude: number | null) and clear mutual-exclusivity documentation.
• Expand descriptions to 100-200 chars. Example: 'Get current weather for a city by name. Returns temperature, wind speed, and conditions. Use when user provides a city name; use get_weather_by_coordinates for lat/long.'
• Document output schemas as JSON Schema. Example for get_weather: {type: 'object', properties: {temperature: {type: number, description: 'Celsius'}, wind_speed: {type: number}, condition: {type: string}}}.
• Add error handling: validate service names in homelab_status against enum [nextcloud, jellyfin, n8n, ollama]; return structured errors like {error: 'Invalid service', valid_options: [...]} instead of mock data.