MCP 服务器

OpenObserve MCP Server

Enables querying logs, traces, and metrics from multiple OpenObserve instances via MCP tools, with parallel execution, batching, and caching.

README

OpenObserve MCP Server

Model Context Protocol (MCP) server for querying OpenObserve instances from AI agents like Claude, Cursor, and OpenCode.

What is MCP?

Model Context Protocol (MCP) is an open protocol that enables AI assistants to connect with external tools and data sources. This server implements MCP to let your AI agent query logs, traces, and metrics from OpenObserve.

Features

Multi-instance support - Query multiple OpenObserve instances in parallel
SQL-based querying - Full PostgreSQL-compatible SQL for logs
Batch execution - Run multiple queries efficiently
LRU caching - Smart caching with TTL for faster results
Pagination - Handle large result sets
MCP stdio transport - Works with any MCP-compatible client

Installation

npm install -g openobserve-mcp

Or use with npx (no install):

npx openobserve-mcp --config /path/to/config.json

Quick Start

Set up authentication (Base64 encoded credentials):

export PROD_GCP_O2_TOKEN=$(echo -n "user@example.com:password" | base64)

Create config file (config.json):

{
  "instances": [{
    "id": "prod-gcp",
    "name": "Production GCP",
    "url": "https://openobserve.example.com",
    "auth": { "type": "env", "envVar": "PROD_GCP_O2_TOKEN" },
    "defaults": { "org": "default", "timeout": 30000, "maxResults": 1000 },
    "capabilities": ["logs", "traces", "metrics"],
    "tags": ["production", "gcp"]
  }]
}

Add to your MCP client (Claude Desktop, Cursor, etc.):

{
  "mcpServers": {
    "openobserve": {
      "command": "npx",
      "args": ["openobserve-mcp", "--config", "/path/to/config.json"],
      "env": { 
        "PROD_GCP_O2_TOKEN": "your_base64_token",
        "PROD_AWS_O2_TOKEN": "your_base64_token"
      }
    }
  }
}

Example MCP server configuration for OpenCode CLI:

{
  "openobserve": {
    "type": "local",
    "command": [
      "npx",
      "openobserve-mcp",
      "--config",
      "/Users/adarsh.ba/breeze/openobserve_mcp/config/openobserve.config.json"
    ],
    "environment": {
      "PROD_GCP_O2_TOKEN": "your_base64_token",
      "PROD_AWS_O2_TOKEN": "your_base64_token"
    },
    "enabled": true
  }
}