---
title: "OpenTelemetry MCP Server Setup Guide and Schema"
slug: "opentelemetry"
category: "observability"
package: "mcp-server-opentelemetry"
transport: "stdio"
canonical: "https://mcp-codex.com/servers/opentelemetry"
source: "MCP Codex (https://mcp-codex.com)"
dateModified: "2026-09-28"
---

# OpenTelemetry MCP Server

> Query distributed traces, spans, and metrics from OpenTelemetry backends.

## Overview
Connect AI agents to OpenTelemetry-compatible backends (Jaeger, Zipkin, Grafana Tempo) for querying distributed traces, analyzing span data, and debugging performance bottlenecks in microservices.

## Quick Install
```bash
npx -y mcp-server-opentelemetry
```

## Environment Variables

| Variable | Description | Required |
| :--- | :--- | :--- |
| `OTEL_ENDPOINT` | OpenTelemetry backend endpoint | **Yes** |

## Client Configurations

### Claude Desktop (`claude_desktop_config.json`)
```json
{
  "mcpServers": {
    "opentelemetry": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-opentelemetry"
      ],
      "env": {
        "OTEL_ENDPOINT": "http://localhost:16686"
      }
    }
  }
}
```

### Cursor (`mcp.json`)
```json
{
  "mcpServers": {
    "opentelemetry": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-server-opentelemetry"
      ],
      "env": {
        "OTEL_ENDPOINT": "http://localhost:16686"
      }
    }
  }
}
```

### OpenAI Codex CLI (`~/.codex/config.toml`)
```toml
[mcp_servers.opentelemetry]
command = "npx"
args = ["-y", "mcp-server-opentelemetry"]

[mcp_servers.opentelemetry.env]
OTEL_ENDPOINT = "http://localhost:16686"
```

## Available Tools (3)

- `query_traces`
- `get_trace`
- `search_spans`

## Documentation & Links
- Canonical Web Page: https://mcp-codex.com/servers/opentelemetry
- Weekly Protocol Changelog: https://mcp-codex.com#newsletter
- GitHub Repository: https://github.com/traceloop/opentelemetry-mcp-server
- Setup Tutorial Guide: https://mcp-codex.com/articles/opentelemetry-tracing-mcp