MCP Catalogs
← Home

otel-server

by cruz-andr·★ 0·Score 36

MCP server for Docker container inspection and OpenTelemetry trace querying via Jaeger with built-in security features.

developer-toolsops-inframonitoring
0
Forks
0
Open issues
2 mo ago
Last commit
2d ago
Indexed

Overview

The otel-server is an intelligent Model Context Protocol server that enables autonomous AI agents to function as Site Reliability Engineers. It bridges static code analysis with runtime observability by allowing LLMs to discover Docker containers, monitor resource utilization, tail logs, and explore OpenTelemetry distributed traces via Jaeger. The server includes robust security features like PII redaction, context window protection, and graceful degradation of API errors.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:AI agents autonomously troubleshooting Docker container issues by checking status and logs
you:Monitoring system resource utilization to detect potential OOM risks before they occur
you:Investigating distributed system issues by querying and visualizing OpenTelemetry traces
you:Does this server require local Docker installation?
you:How does the server handle sensitive data?

When to choose this

Choose this server if you're managing Docker containers and distributed systems with OpenTelemetry and need an AI assistant to help with observability and troubleshooting.

When NOT to choose this

Don't choose this if you don't use Docker or OpenTelemetry, as its functionality is limited to these ecosystems.

Tools this server exposes

6 tools extracted from the README
  • list_containers

    Discovers local Docker containers by name or status.

  • get_container_status

    Returns container state, health status, exit codes, and restart loops.

  • get_container_stats

    Captures CPU, Memory, and I/O snapshots with warnings for high memory usage.

  • fetch_logs

    Tails container logs with ANSI stripping, grep filtering, and time-range navigation.

  • search_error_traces

    Queries Jaeger for traces flagged with errors within a specific timeframe.

  • get_trace_tree

    Recursively builds and formats a distributed trace into a readable Markdown list.

Comparable tools

docker-mcpjaeger-uigrafana-mcp

Installation

Installation

Prerequisites

  • Node.js v22 or higher
  • Docker (running locally)
  • Jaeger (optional, running locally or accessible via HTTP API)

Build Instructions

git clone https://github.com/cruz-andr/otel-server.git
cd otel-server
npm install
npm run build

Configuration

Configure via environment variables by copying .env.example to .env or passing them through your MCP client.

Claude Desktop Configuration

{
  "mcpServers": {
    "otel-server": {
      "command": "node",
      "args": ["/absolute/path/to/otel-server/build/index.js"]
    }
  }
}

Hosted Server

claude mcp add --transport http otel-server https://otel-server.fly.dev/mcp

FAQ

Does this server require local Docker installation?
Yes, the server needs Docker running locally with socket access (default at /var/run/docker.sock).
How does the server handle sensitive data?
The server automatically redacts PII and credentials including JWTs, Bearer tokens, emails, and credit cards before returning data to the LLM.

Compare otel-server with

otel-server vs mcp-server-chartotel-server vs everythingotel-server vs filesystemotel-server vs timeotel-server vs sequentialthinking
GitHub →

Last updated · Auto-generated from public README + GitHub signals.