Best for
- SRE and development teams already running LogClaw or using its cloud service
- Engineers who want incident and log data directly in Claude Code, Cursor, or Windsurf
L
LogClaw MCP Server
Official
Connect AI coding tools to LogClaw incidents, logs, and anomaly detections
GitHub source repository ↗
56FMRS · C
Official LogClaw MCP server exposing 8 tools over stdio so AI coding tools can access incidents, logs, and anomaly data. The source states the tool count but does not name the individual tools, so the tool list is left empty. A solid fit for teams already on LogClaw.
The LogClaw MCP Server is the official MCP server for LogClaw, a log intelligence platform offering OTLP-native log ingestion, anomaly detection, trace correlation, and AI root cause analysis. It runs over stdio, is published on npm as logclaw-mcp-server (version 1.0.4), and exposes 8 tools that let MCP-compatible clients such as Claude Code, Cursor, and Windsurf query LogClaw incidents, logs, and anomaly detections directly. A hosted remote version is also available at https://mcp.logclaw.ai (OAuth 2.1).
Setup
- Run via npx logclaw-mcp-server, or from the repo's apps/logclaw-mcp-server directory run npm install && npm run build and start it. 2. Set the required environment variable LOGCLAW_API_KEY (a LogClaw project API key starting with lc_proj_). 3. Optionally set LOGCLAW_ENDPOINT to your LogClaw auth proxy endpoint (default https://ticket.logclaw.ai). 4. Configure the server in your MCP client using the stdio transport.
Fit and risk
Not for
- Users who do not use the LogClaw platform
- Anyone without a valid LOGCLAW_API_KEY (lc_proj_ prefix)
Required permissions
- Requires a valid LogClaw project API key (lc_proj_ prefix)
- Network access to the LogClaw auth proxy endpoint (default https://ticket.logclaw.ai)
Risks and side effects
- API key leakage could expose project incident and log data; secure environment variables accordingly
- Logs exposed to AI tools may contain sensitive operational information
- Depends on remote endpoint availability; the self-hosted local stack runs single-tenant with AI RCA disabled by default
Troubleshooting
- Verify LOGCLAW_API_KEY is set and starts with lc_proj_
- If self-hosting or using a custom endpoint, check LOGCLAW_ENDPOINT points to the correct auth proxy
- Configure the client with stdio transport and inspect its logs
- Confirm the LogClaw backend services are reachable
Use cases
Let AI coding assistants query active incidents and their severities in LogClaw
Retrieve and analyze logs and anomaly detections inside your IDE
Bring incident context into code-fixing workflows
Supported clients
Claude CodeFull support
CursorFull support
WindsurfFull support