RCA Connector

⚠️ Pre-release. RCA-MCP is still in active development and testing. The private API backend (api.rca-mcp.com) is not yet deployed, so this connector cannot be used end-to-end yet. Tool schemas and documentation below reflect the current development snapshot and may change before the first public launch. Watch this repo or check rca-mcp.com for launch updates.

What is RCA-MCP?

The only MCP server purpose-built for causal Root Cause Analysis. 55 tools covering causal graph construction, 13 RCA model families (including Salesforce PyRCA algorithms), multi-model consensus, and PDF/HTML/Excel/Markdown report generation. Works with Claude, Ollama, Groq, OpenAI, Gemini, LangChain — 9 providers.

---

Quick Start (2 minutes)

git clone https://github.com/dave1362/rca-mcp-connector.git
cd rca-mcp-connector
pip install -r requirements.txt
cp .env.example .env
# Edit .env: set RCA_MCP_API_KEY to your key from https://rca-mcp.com/upgrade

Get an API key at rca-mcp.com — a free tier is available, no credit card required.

---

Claude Code Setup

Add to .mcp.json in your workspace root:

{
  "mcpServers": {
    "rca-mcp": {
      "command": "python",
      "args": ["/absolute/path/to/rca-mcp-connector/connector/server.py"],
      "env": {
        "RCA_MCP_API_URL": "https://api.rca-mcp.com",
        "RCA_MCP_API_KEY": "your_api_key_here"
      }
    }
  }
}

Ollama Setup

go install github.com/mark3labs/mcphost@latest
mcphost -m ollama:qwen3:14b --config providers/mcp-servers.json

OpenAI Agents SDK

from agents import Agent, MCPServerStdio
import asyncio

async def main():
    async with MCPServerStdio(
        params={
            "command": "python",
            "args": ["connector/server.py"],
            "env": {"RCA_MCP_API_KEY": "your_key"},
        }
    ) as rca_server:
        agent = Agent(name="RCA Agent", model="gpt-4o", mcp_servers=[rca_server])
        result = await agent.run("Find the root cause of the API latency spike.")
        print(result.final_output)

asyncio.run(main())

LangChain

from langchain_mcp_adapters.client import MultiServerMCPClient
from langchain_anthropic import ChatAnthropic
from langgraph.prebuilt import create_react_agent
import asyncio

async def main():
    async with MultiServerMCPClient({
        "rca-mcp": {
            "command": "python", "args": ["connector/server.py"],
            "env": {"RCA_MCP_API_KEY": "your_key"}, "transport": "stdio",
        }
    }) as client:
        tools = await client.get_tools()
        agent = create_react_agent(ChatAnthropic(model="claude-sonnet-4-6"), tools)
        result = await agent.ainvoke({"messages": [{"role": "user", "content": "Run an FMEA analysis"}]})
        print(result["messages"][-1].content)

asyncio.run(main())

See providers/ for ready-to-use config templates and full examples (Groq, Gemini, OpenRouter, Claude Desktop).

---

All 55 Tools

See rca-mcp.com/pricing for full tier details and limits.

Legend: ✅ All plans · 🌟 Starter+ · 💎 Pro+ · 👑 Enterprise only

Group A — Causal Graph

Tool	Plan	Description
rca_graph_create	✅	Create an empty causal DAG
rca_graph_get	✅	Retrieve a graph as JSON / DOT / adjacency
rca_graph_score	✅	Structural quality scoring
rca_graph_discover	💎	Auto-discover causal structure from data (PC-algorithm)
rca_graph_merge	💎	Merge two graphs for cross-system RCA
rca_graph_list_versions	🌟	List historical graph snapshots
rca_graph_restore_version	👑	Roll back to a historical version (admin)
rca_graph_delete	👑	Delete a graph permanently (admin)

Group B — RCA Models

Tool	Plan	Description
rca_model_create	✅	Register a model spec (up to plan's model limit)
rca_model_list	✅	List registered models
rca_model_update_status	✅	Advance model lifecycle
rca_model_validate	✅	Validate on hold-out data
rca_model_delete	👑	Delete a model (admin)

Group C — Analysis

Tool	Plan	Description
rca_analysis_run	✅	Run RCA, get ranked root causes
rca_analysis_run_async	💎	Submit a long-running analysis (returns task_id)
rca_analysis_poll_task	💎	Poll an async task's status
rca_analysis_ensemble	💎	Combine 2-5 models via weighted-vote consensus
rca_analysis_get_result	✅	Retrieve a stored result
rca_analysis_list_results	✅	Paginated result listing
rca_analysis_query_results	✅	Filter results by family/confidence/tags
rca_analysis_compare	✅	Cross-model consensus comparison
rca_analysis_explain	✅	Narrative explanation of a result
rca_analysis_batch	🌟	Batch analysis (5 incidents Starter, 20 Pro+)

Group D — Graph Operations

Tool	Plan	Description
rca_graph_add_node	✅	Add a typed node
rca_graph_remove_node	✅	Remove a node
rca_graph_add_edge	✅	Add a directed causal edge (cycle-safe)
rca_graph_remove_edge	✅	Remove an edge
rca_graph_score_paths	✅	Rank causal paths to a target node
rca_graph_markov_blanket	✅	Get parents/children/co-parents of a node

Group E — Admin & Auth

Tool	Plan	Description
rca_auth_generate_token	✅	Generate API key + JWT — call this first
rca_auth_revoke_token	👑	Invalidate a JWT before expiry (admin)
rca_auth_list_keys	👑	List API keys (admin + Pro+ multi-key feature)
rca_auth_rotate_key	👑	Rotate an API key (admin + Pro+ multi-key feature)
rca_admin_health	✅	Server health snapshot
rca_admin_read_audit_log	✅	Read structured audit entries
rca_admin_purge_namespace	👑	Delete all records in a namespace (admin)
rca_admin_show_plan_info	✅	Show current plan, limits, and upgrade options

Group F — Reports & Providers

Tool	Plan	Description
rca_report_generate	✅/🌟/💎	Markdown all plans; HTML/PDF Starter+; Excel Pro+
rca_report_compare	✅/🌟	Comparative report; Markdown all, HTML Starter+
rca_provider_list_configs	✅	Setup instructions for any of the 9 providers

Group G — PyRCA Algorithms [PyRCA — BSD-3-Clause]

Tool	Plan	Description
rca_pyrca_epsilon_diagnosis	🌟	z-score anomalous metric identification
rca_pyrca_random_walk	🌟	Personalised PageRank root cause localisation
rca_pyrca_ht_diagnosis	🌟	Hypothesis testing with descendant adjustment
rca_pyrca_validate_setup	✅	Check PyRCA strategy and licence compliance

Group H — Equipment Knowledge (Phase 9)

Tool	Plan	Description
rca_guide_ingest	🌟	Ingest a troubleshooting guide (markdown/plain/json_dtree)
rca_guide_search	✅	Search guides by symptom (TF-IDF); Free capped at 3 results
rca_guide_get	✅	Retrieve a full guide or a specific section
rca_guide_list	✅	List ingested guides with equipment_type/tag filters
rca_dtree_start	🌟	Begin an interactive diagnostic decision-tree session
rca_dtree_answer	🌟	Answer a diagnostic question; returns next question or diagnosis
rca_dtree_list_sessions	🌟	List diagnostic sessions with equipment/status filters
rca_guide_generate_report	🌟/💎	Maintenance report from a resolved session; Markdown Starter+, PDF/HTML Pro+
rca_dtree_generate_from_fmea	🌟	Auto-generate a decision tree from FMEA results
rca_guide_pdf_preview	🌟	Preview PDF quality before ingestion
rca_guide_ingest_pdf	🌟	Parse a PDF equipment manual → knowledge base

Ships with 4 built-in sample guides (pump, vacuum interface valve, ML pipeline, CFD solver) auto-loaded on first startup, and supports 26 equipment types. PDF ingestion supports 4 parsing strategies (text_native, ocr, table, mixed).

---

Get an API Key

rca-mcp.com/upgrade — Free tier available, plans from $19/month (Starter), $79/month (Pro), $399/month (Enterprise).

---

Third-Party Licences

PyRCA (Salesforce): BSD-3-Clause Copyright (c) 2022, salesforce.com, inc. https://github.com/salesforce/PyRCA

Algorithms in rca_pyrca_* tools are independently-written adaptations of PyRCA's published methods (Zheng et al. 2023, arXiv:2306.11417), not direct copies of PyRCA source code, per the private API's models/pyrca_adapter.py.

---

Citing RCA-MCP

@software{rcamcp2026,
  title  = {RCA-MCP: An MCP Server for Causal Root Cause Analysis},
  author = {davetj},
  year   = {2026},
  url    = {https://github.com/dave1362/rca-mcp-connector},
  note   = {v3.0.0}
}