# LogClaw MCP server

> Connect AI coding tools to LogClaw incidents, logs, and anomaly detections.

- Repository: https://github.com/logclaw/logclaw
- Page: https://hostedmcp.org/servers/logclaw/logclaw
- Transport: stdio
- Stars: 79
- Latest version: 1.0.4

## Install

### Claude Code

```sh
claude mcp add logclaw -- npx -y logclaw-mcp-server
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "logclaw": {
      "command": "npx",
      "args": [
        "-y",
        "logclaw-mcp-server"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "logclaw": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "logclaw-mcp-server"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"logclaw","type":"stdio","command":"npx","args":["-y","logclaw-mcp-server"]}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "logclaw": {
      "command": "npx",
      "args": [
        "-y",
        "logclaw-mcp-server"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add logclaw -- npx -y logclaw-mcp-server
```

### Codex config.toml (`~/.codex/config.toml`)

```toml
[mcp_servers.logclaw]
command = "npx"
args = ["-y", "logclaw-mcp-server"]
```

### Gemini CLI

```sh
gemini mcp add -s user logclaw npx -- -y logclaw-mcp-server
```

### Gemini settings.json (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "logclaw": {
      "command": "npx",
      "args": [
        "-y",
        "logclaw-mcp-server"
      ]
    }
  }
}
```

## Environment variables

| Name | Required | Secret | Description |
|---|---|---|---|
| `KAFKA_BROKERS` | no | no |  |
| `LOGCLAW_API_KEY` | no | no | Your LogClaw project API key (starts with lc_proj_) |
| `LOGCLAW_ENDPOINT` | no | no | LogClaw auth proxy endpoint (default: https://ticket.logclaw.ai) |
| `OPENSEARCH_ENDPOINT` | no | no |  |

## FAQ

### Does the local Docker Compose stack support AI root cause analysis?

No, the local stack runs in single-tenant mode with LLM-powered root cause analysis disabled. You must use the managed cloud or deploy to Kubernetes with an LLM provider specified to use AI RCA.

### What protocols does LogClaw use for log ingestion?

LogClaw ingests logs using the OpenTelemetry Protocol (OTLP), which is the industry standard.
