otel-mcp-server

shiftyp/otel-mcp-server

3.3

If you are the rightful owner of otel-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

This MCP server enables natural language interactions with OpenTelemetry data, transforming observability data into actionable intelligence.

The AI-Powered OpenTelemetry Analysis server bridges the gap between AI assistants and OpenTelemetry data, allowing users to interact with their observability stack using natural language. By leveraging the Model Context Protocol (MCP), this server provides a structured interface to OpenTelemetry data stored in Elasticsearch/OpenSearch. Users can ask questions in plain English, and the AI will understand the intent, construct appropriate queries, analyze results, and present insights. This approach eliminates the need for complex query languages and manual data correlation, making it easier for teams to extract insights from their telemetry data. The server supports various use cases, including incident response, performance analysis, and anomaly detection, democratizing access to observability insights across different team roles.

Features

  • Natural Language Queries: Interact with telemetry data using plain English instead of complex query languages.
  • Cross-Signal Correlation: Automatically correlate traces, metrics, and logs for comprehensive insights.
  • Pattern Recognition: Discover anomalies and trends that might be missed manually.
  • Contextual Understanding: AI understands service relationships and dependencies for more accurate analysis.
  • Vendor Neutrality: Works with Elasticsearch, OpenSearch, Jaeger, Prometheus, and more.

Usages

usage with windsurf claude desktop

{
  "mcpServers": {
    "otel-mcp-server": {
      "command": "npx",
      "args": ["-y", "otel-mcp-server"],
      "env": {
        "OPENSEARCH_URL": "http://localhost:9200",
        "USERNAME": "elastic",
        "PASSWORD": "changeme",
        "OPENAI_API_KEY": "sk-..."  // Optional: for ML-powered features
      }
    }
  }
}