Skip to main content
Glama

πŸ€– Elastic Agent Builder Integration

Connect to Elastic Agent Builder for AI-powered application health monitoring using OTEL data from Elastic Serverless Observability.

✨ Features

  • AI-Powered Analysis through Elastic Agent Builder

  • Real-time Insights from OTEL metrics, traces, and logs

  • Built-in Tools for comprehensive monitoring

  • Easy Integration with Cursor IDE and other MCP clients

  • No Local Server Required - direct connection to Elastic

Related MCP server: otel-mcp-server

🎯 Available Agent Builder Tools

  • platform.core.search - Search OTEL data in Elasticsearch

  • platform.core.execute_esql - Execute ESQL queries on OTEL data

  • platform.core.generate_esql - Generate ESQL queries for analysis

  • Built-in Analysis - Error patterns, performance metrics, resource utilization

  • Real-time Monitoring - Application health, service dependencies, bottlenecks

πŸš€ Quick Start

Prerequisites

  • Elastic Serverless Observability endpoint with Agent Builder enabled

  • Valid API key for Elastic

  • Cursor IDE (optional, for MCP integration)

Installation

  1. Clone the repository

    git clone git@github.com:poulsbopete/o11y-mcp-server.git
    cd o11y-mcp-server
  2. Configure credentials

    # Option 1: Use setup script (recommended)
    ./setup-env.sh
    # Then edit .env file with your credentials
    
    # Option 2: Set environment variables
    export ELASTIC_ENDPOINT=https://your-endpoint.kb.us-east-1.aws.elastic.cloud
    export ELASTIC_API_KEY=your-api-key
  3. Connect to Agent Builder

    # Test Agent Builder connection
    ./connect-to-agent-builder.sh
    
    # Query application health
    ./query-application-health.sh
  4. Set up Cursor integration (optional)

    # Create Cursor MCP configuration
    ./setup-cursor-agent-builder.sh

πŸ”— Cursor Integration

Quick Setup

./setup-cursor-agent-builder.sh

Manual Setup

  1. Run the Cursor configuration script

  2. Restart Cursor IDE

  3. Connect to Elastic Agent Builder

  4. Ask questions like "What's the health of my applications?"

πŸ“Š Data Sources

Agent Builder connects to Elastic Serverless Observability and analyzes:

  • Metrics: metrics-* indices (CPU, memory, custom metrics)

  • Logs: logs-* indices (application logs, errors, events)

  • Traces: traces-* indices (distributed tracing data)

πŸ› οΈ Configuration

Environment Variables

  • ELASTIC_ENDPOINT - Your Elastic Serverless endpoint

  • ELASTIC_API_KEY - Your Elastic API key

  • LOG_LEVEL - Logging level (default: INFO)

Cursor MCP Configuration

The setup script creates ~/.cursor/mcp_config.json with Agent Builder integration.

πŸ“ Example Queries

Once connected to Agent Builder, you can ask:

  • "What's the overall health of my applications?"

  • "Show me metrics for the payment service"

  • "What errors are happening in the last hour?"

  • "Find the slowest operations in my system"

  • "Analyze resource utilization across all services"

  • "Show me trace analysis for the checkout flow"

  • "Generate an ESQL query for error analysis"

πŸ”§ Troubleshooting

Agent Builder Connection Issues

  • Verify Elastic endpoint and API key

  • Check if Agent Builder is enabled on your Elastic instance

  • Test connection: ./connect-to-agent-builder.sh

No Data Found

  • Check if applications are sending OTEL data

  • Verify time range includes recent data

  • Ensure proper Elastic cluster configuration

Cursor Integration Issues

  • Verify MCP configuration in Cursor settings

  • Check Agent Builder is accessible

  • Restart Cursor after configuration changes

πŸ“š Documentation

  • CURSOR-MCP-INTEGRATION.md - Cursor integration guide

  • connect-to-agent-builder.sh - Agent Builder connection script

  • query-application-health.sh - Application health analysis script

🀝 Contributing

  1. Fork the repository

  2. Create a feature branch: git checkout -b feature-name

  3. Commit changes: git commit -am 'Add feature'

  4. Push to branch: git push origin feature-name

  5. Submit a pull request

πŸ“„ License

This project is licensed under the MIT License - see the LICENSE file for details.

πŸ™ Acknowledgments


Ready to monitor your application health with AI-powered insights through Elastic Agent Builder! πŸŽ‰

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Enables LLMs to query telemetry data via the Spyglass AI agent, providing intelligent insights about application performance, errors, and bottlenecks.
      1
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables natural language querying and analysis of OpenTelemetry traces, metrics, and logs stored in Elasticsearch/OpenSearch, allowing AI assistants to investigate performance issues, find root causes, and explore system behavior.
      14
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Connects AI assistants to New Relic with read-only access to NRQL, logs, metrics, traces, alerts, and more. Offers optional, gated write operations with a dry-run and confirmation workflow.
      49 npm
      Apache 2.0