logclaw-mcp-server — Claude Code MCP server: incidents and logs
Connects AI agents to LogClaw incidents, logs, and anomaly detections. Works with cloud API or self-hosted instances.
claude mcp add --env LOGCLAW_API_KEY=<LOGCLAW_API_KEY> --env LOGCLAW_ENDPOINT=<LOGCLAW_ENDPOINT> --transport stdio logclaw-mcp-server -- npx -y logclaw-mcp-serverWhat logclaw-mcp-server does
The server surfaces LogClaw incidents in agent context — a log monitoring platform with anomaly detection and AI root cause analysis. The agent sees traces, severity, blast radius, and event correlations.
LogClaw deploys in three modes: cloud SaaS (console.logclaw.ai), Docker Compose for a local stack without Kubernetes, and a full Helm chart for production. The server requires two environment variables: LOGCLAW_API_KEY (project key) and LOGCLAW_ENDPOINT (API address — cloud or your instance URL). Without them it starts but cannot read data.
In self-hosted mode, AI root cause analysis runs through Ollama, OpenAI, or Claude — set via LLM_PROVIDER. Cloud version includes RCA out of the box.
What it helps with
- read incident list with severity and timestamps
- find anomalies in logs for a time range
- get trace correlations for a specific incident
- check event blast radius across related services
Example prompts for the agent
- Show LogClaw incidents from the last hour
- Find anomalies in
auth-servicelogs since 2 PM - Pull traces for incident
INC-2341and explain where it broke
How to install logclaw-mcp-server
Platform CLI
claude mcp add --env LOGCLAW_API_KEY=<LOGCLAW_API_KEY> --env LOGCLAW_ENDPOINT=<LOGCLAW_ENDPOINT> --transport stdio logclaw-mcp-server -- npx -y logclaw-mcp-serverReplace <…> with your own values
Platform CLI
codex mcp add logclaw-mcp-server --env LOGCLAW_API_KEY=<LOGCLAW_API_KEY> --env LOGCLAW_ENDPOINT=<LOGCLAW_ENDPOINT> -- npx -y logclaw-mcp-serverReplace <…> with your own values
Config file
{
"mcpServers": {
"logclaw-mcp-server": {
"command": "npx",
"args": [
"-y",
"logclaw-mcp-server"
],
"env": {
"LOGCLAW_API_KEY": "<LOGCLAW_API_KEY>",
"LOGCLAW_ENDPOINT": "<LOGCLAW_ENDPOINT>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"logclaw-mcp-server": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"logclaw-mcp-server"
],
"env": {
"LOGCLAW_API_KEY": "<LOGCLAW_API_KEY>",
"LOGCLAW_ENDPOINT": "<LOGCLAW_ENDPOINT>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e LOGCLAW_API_KEY=<LOGCLAW_API_KEY> -e LOGCLAW_ENDPOINT=<LOGCLAW_ENDPOINT> logclaw-mcp-server npx -y logclaw-mcp-serverReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about logclaw-mcp-server
How to install logclaw-mcp-server in Claude Code?
Run `claude mcp add --env LOGCLAW_API_KEY=<LOGCLAW_API_KEY> --env LOGCLAW_ENDPOINT=<LOGCLAW_ENDPOINT> --transport stdio logclaw-mcp-server -- npx -y logclaw-mcp-server`. Get the key and endpoint from your LogClaw project settings.
What does logclaw-mcp-server do?
Connects agents to LogClaw incidents, logs, and anomaly detections. The agent reads traces, event correlations, and AI root cause analysis.
Does it work with self-hosted LogClaw?
Yes. Point LOGCLAW_ENDPOINT to your instance — cloud console.logclaw.ai or local http://localhost:3000. Create the API key in project settings.
