nitishagar/OTel-query-mcp

OpenTelemetry Query Server

An MCP (Model Context Protocol) server that provides tools for querying observability data from various backends including OpenTelemetry Collector, Grafana, Elastic Cloud, and OpenSearch.

Features

• 🔍 Multi-Backend Support: Query traces, logs, and metrics from multiple observability platforms
• 🚀 Fast & Async: Built with async Python for high performance
• 🔧 MCP Compatible: Works with any MCP-enabled client
• 💾 Smart Caching: LRU cache with configurable TTL per data type
• 🛡️ Type Safe: Full type annotations with Pydantic models
• 📊 Rich Query Tools: Search traces, logs, metrics, and correlate across signals
• ⚡ Production Ready: Structured logging, retries, health checks, and graceful shutdown

Quick Start

Installation

# From source
git clone https://github.com/nitishagar/OTel-query-mcp.git
cd OTel-query-mcp
pip install -e .

# From PyPI (when published)
pip install otel-query-server

Basic Usage

1. Create configuration (config.yaml):

server:
  name: my-otel-server
  log_level: INFO

backends:
  otel_collector:
    enabled: true
    endpoint: localhost:4317
    insecure: true

2. Run the server:

# Using Python module
python -m otel_query_server.server --config config.yaml

# Using FastMCP CLI
fastmcp run otel_query_server.server:mcp

# With environment variables
export OTEL_QUERY_BACKENDS__OTEL_COLLECTOR__ENDPOINT=prod:4317
python -m otel_query_server.server

3. Connect with MCP client:

mcp-client connect stdio -- python -m otel_query_server.server

Local Development

Prerequisites

• Python 3.9+
• Docker (optional, for backend services)
• Git

Setup

# Clone and setup
git clone https://github.com/nitishagar/OTel-query-mcp.git
cd OTel-query-mcp

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

# Install with dev dependencies
pip install -e ".[dev,test]"

# Run tests
pytest --cov=otel_query_server

# Start backend services
docker-compose -f examples/docker-compose.yaml up -d

Configuration

The server supports configuration via:

1. YAML configuration file
2. Environment variables (prefix: OTEL_QUERY_)
3. Command-line arguments

Key configuration options:

# Server settings
server:
  log_level: DEBUG  # DEBUG, INFO, WARNING, ERROR
  max_concurrent_requests: 10

# Caching
cache:
  enabled: true
  max_size: 1000
  trace_ttl_seconds: 600  # 10 minutes
  log_ttl_seconds: 300    # 5 minutes
  metric_ttl_seconds: 60  # 1 minute

# Backends (see examples/config.yaml for full options)
backends:
  otel_collector:
    endpoint: localhost:4317
    insecure: true
  grafana:
    tempo_url: http://localhost:3200
    loki_url: http://localhost:3100
  # ... more backends

See for detailed setup instructions.

Supported Backends

• OpenTelemetry Collector (OTLP gRPC/HTTP)
• Grafana Stack:
  • Tempo (traces)
  • Loki (logs)
  • Prometheus (metrics)
• Elastic Cloud (APM, Elasticsearch)
• OpenSearch

Available MCP Tools

• search_traces: Search for distributed traces
• search_logs: Query application logs
• query_metrics: Retrieve and aggregate metrics
• get_service_health: Check service health status
• correlate_trace_logs: Find logs related to a trace
• get_server_info: Get server configuration info

CI/CD

This project uses GitHub Actions for continuous integration:

• Automated Testing: Tests run on every PR across Python 3.9-3.12 on Linux, macOS, and Windows
• Code Quality: Black formatting, Ruff linting, MyPy type checking
• Security Scanning: Trivy and Bandit security analysis
• Coverage Reports: Automatic upload to Codecov
• Docker Builds: Containerized builds on every commit

See for the complete pipeline.

Documentation

• - Detailed setup and configuration
• - MCP tools reference
• - System design and patterns

Contributing

We welcome contributions! Please see for guidelines.

License

Apache License 2.0 - see for details.