by dave1362
RCA MCP is an MCP server for causal root cause analysis, with 56 tools for causal graphs, RCA models and report generation. The rca-mcp-connector package connects MCP clients to its hosted backend.
PluginContact for pricing
Step 1: Get a free API key from the service named in the README.
Step 2: Add an rca-mcp server entry to your MCP client config, for example .mcp.json in Claude Code, with command uvx and args rca-mcp-connector.
Step 3: Set RCA_MCP_API_URL to the backend URL given in the README and RCA_MCP_API_KEY to your key.
Step 4: Restart the client so it loads the server.
Step 5: Ask the agent to find the root cause of an issue, run an analysis such as FMEA, or produce a report.
Run it with uvx rca-mcp-connector, or install the PyPI package with pip install rca-mcp-connector. No clone is needed.
Yes. The README says to get a free key, with no credit card required, and set it as RCA_MCP_API_KEY.
The README gives setup for Claude Code, Ollama, the OpenAI Agents SDK and LangChain, and says it works with 10 providers.
Requirements
CodeSampleX MCPCodeSampleX (csx) gives AI coding agents a memory of which package versions and APIs actually built or failed, and in which environment. Its MCP server is part of the csx binary.
World Model MCPWorld Model MCP is a local MCP server from PuterVision that keeps a persistent 3D and 2D spatial model of entities for AI agents. It tracks objects, simulates movement and checks collisions without any LLM in the loop.