OpenTelemetry
Unified MCP server for querying OpenTelemetry traces across multiple backends (Jaeger, Tempo, Traceloop, etc.), enabling AI agents to analyze distributed traces for automated debugging and observability.
What is OpenTelemetry?
OpenTelemetry is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to unified mcp server for querying opentelemetry traces across multiple backends (jaeger, tempo, traceloop, etc.), enabling ai agents to analyze distributed traces for automated debugging and observabili...
Unified MCP server for querying OpenTelemetry traces across multiple backends (Jaeger, Tempo, Traceloop, etc.), enabling AI agents to analyze distributed traces for automated debugging and observability.
This server falls under the Monitoring & Observability category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Unified MCP server for querying OpenTelemetry traces across
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx opentelemetryConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use OpenTelemetry
The OpenTelemetry MCP server lets AI agents query distributed traces from multiple observability backends — Jaeger, Grafana Tempo, and Traceloop — through a unified interface, enabling natural language debugging of microservices, latency analysis, and LLM token usage tracking. It exposes 10 specialized tools covering trace search, span analysis, error detection, and LLM-specific metrics, so engineers can ask questions like 'why is the checkout service slow today?' and get grounded answers from real trace data without switching to a separate dashboard.
Prerequisites
- Python 3.10 or newer with pipx or uvx installed
- A running OpenTelemetry-compatible backend: Jaeger (port 16686), Grafana Tempo, or a Traceloop account with API key
- Services already instrumented with OpenTelemetry SDKs sending traces to your backend
- Claude Desktop or another MCP-capable client
Confirm your observability backend is accessible
Verify that your Jaeger, Tempo, or Traceloop backend is reachable from the machine running the MCP server. For local Jaeger, the default UI and API URL is http://localhost:16686.
curl http://localhost:16686/api/servicesTest the server with pipx
Run the server once with pipx to confirm connectivity to your backend before adding it to your MCP client.
pipx run opentelemetry-mcp --backend jaeger --url http://localhost:16686Add the server to your MCP client configuration using environment variables
Edit claude_desktop_config.json to launch the OpenTelemetry MCP server with your backend type and URL. For Traceloop, also add BACKEND_API_KEY.
{
"mcpServers": {
"opentelemetry": {
"command": "uvx",
"args": ["opentelemetry-mcp"],
"env": {
"BACKEND_TYPE": "jaeger",
"BACKEND_URL": "http://localhost:16686",
"LOG_LEVEL": "INFO",
"MAX_TRACES_PER_QUERY": "100"
}
}
}
}For Traceloop backend, add the API key
If using Traceloop as your backend, add BACKEND_API_KEY to the env section with your Traceloop API key.
{
"mcpServers": {
"opentelemetry": {
"command": "uvx",
"args": ["opentelemetry-mcp"],
"env": {
"BACKEND_TYPE": "traceloop",
"BACKEND_URL": "https://api.traceloop.com",
"BACKEND_API_KEY": "your_traceloop_api_key"
}
}
}
}Restart your MCP client and start debugging
Restart Claude Desktop, then ask it to list services in your traces to confirm the connection is working.
OpenTelemetry Examples
Client configuration
Complete claude_desktop_config.json entry for the OpenTelemetry MCP server connecting to a local Jaeger instance.
{
"mcpServers": {
"opentelemetry": {
"command": "uvx",
"args": ["opentelemetry-mcp"],
"env": {
"BACKEND_TYPE": "jaeger",
"BACKEND_URL": "http://localhost:16686",
"MAX_TRACES_PER_QUERY": "100",
"BACKEND_TIMEOUT": "30"
}
}
}
}Prompts to try
These prompts use the 10 tools exposed by the OpenTelemetry MCP server for trace analysis, error detection, and LLM usage monitoring.
- "List all services that have sent traces in the last hour"
- "Show me traces with errors from the checkout service in the last 30 minutes"
- "Find the slowest traces for the payment service today"
- "How many tokens did we use for each LLM model in the last 24 hours?"
- "Which requests used the most tokens today?"
- "Compare the performance of GPT-4 vs Claude across our traces this week"Troubleshooting OpenTelemetry
Connection refused or timeout when querying the backend
Verify the BACKEND_URL is accessible from the machine running the MCP server. For Jaeger running in Docker, use the host machine's IP or Docker bridge IP instead of localhost if the MCP server runs outside Docker. Increase BACKEND_TIMEOUT if the backend is slow to respond.
list_services returns an empty list
No services appear if no traces have been ingested recently. Confirm your applications are instrumented with OpenTelemetry and are actively sending traces to the backend. Generate some test traffic, then retry.
uvx command not found
Install uv with 'curl -LsSf https://astral.sh/uv/install.sh | sh', which includes uvx. Alternatively, install with pip: 'pip install uv'. Then open a new shell before running the MCP client.
Frequently Asked Questions about OpenTelemetry
What is OpenTelemetry?
OpenTelemetry is a Model Context Protocol (MCP) server that unified mcp server for querying opentelemetry traces across multiple backends (jaeger, tempo, traceloop, etc.), enabling ai agents to analyze distributed traces for automated debugging and observability. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install OpenTelemetry?
Follow the installation instructions on the OpenTelemetry GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with OpenTelemetry?
OpenTelemetry works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is OpenTelemetry free to use?
Yes, OpenTelemetry is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
OpenTelemetry Alternatives — Similar Monitoring & Observability Servers
Looking for alternatives to OpenTelemetry? Here are other popular monitoring & observability servers you can use with Claude, Cursor, and VS Code.
Netdata
★ 78.9kReal-time infrastructure monitoring with metrics, logs, alerts, and ML-based anomaly detection.
Kubeshark
★ 11.9keBPF-powered network observability for Kubernetes. Indexes L4/L7 traffic with full K8s context, decrypts TLS without keys. Queryable by AI agents via MCP and humans via dashboard.
Mission Control
★ 4.9kSelf-hosted AI agent orchestration platform: dispatch tasks, run multi-agent workflows, monitor spend, and govern operations from one mission control dashboard.
Grafana
★ 3.0kThis MCP server enables natural-language querying of Grafana logs by automatically detecting log sources and service labels. It provides read-only access to log data with intelligent caching for efficient repeat queries.
Sentrux
★ 2.4kReal-time architectural sensor that helps AI agents close the feedback loop, enabling recursive self-improvement of code quality. Pure Rust.
OpenInference
★ 986OpenTelemetry Instrumentation for AI Observability
Browse More Monitoring & Observability MCP Servers
Explore all monitoring & observability servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up OpenTelemetry in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use OpenTelemetry?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.