Skip to main content

What is Heimdall?

Heimdall is a comprehensive observability platform designed for monitoring Model Context Protocol (MCP) servers and AI/LLM applications. Built on OpenTelemetry standards, it provides real-time tracing, metrics, and insights into your AI infrastructure. Heimdall Dashboard

Key Features

Real-time Tracing

Track every tool call, resource access, and prompt execution in your MCP servers with detailed span information.

Dashboard Analytics

Visualize latency, error rates, and usage patterns with an intuitive dashboard interface.

Easy Integration

Simple SDK decorators for Python and JavaScript/TypeScript. Just wrap your functions and start tracing.

Self-hosted

Run entirely on your own infrastructure. Your data never leaves your servers.

OpenTelemetry Native

Heimdall is built on industry-standard OpenTelemetry protocols. This means:
  • Compatibility: Works with any OTLP-compatible tooling
  • Future-proof: Based on CNCF-backed observability standards
  • Extensible: Easy to integrate with existing observability stacks

Architecture

Heimdall uses a simple but powerful architecture:
┌─────────────────┐     OTLP/HTTP      ┌─────────────────┐
│   MCP Server    │ ──────────────────▶│  Heimdall       │
│   + SDK         │    Port 4318       │  Backend        │
└─────────────────┘                    └────────┬────────┘

                                                │ REST API

                                       ┌────────▼────────┐
                                       │  Heimdall       │
                                       │  Frontend       │
                                       │  Port 5173      │
                                       └─────────────────┘
  1. Your MCP Server: Instrumented with the Heimdall SDK (Python or JavaScript)
  2. Heimdall Backend: Receives OTLP traces and stores them
  3. Heimdall Frontend: Beautiful dashboard for viewing and analyzing traces

SDKs

Heimdall provides official SDKs for instrumenting your MCP servers:
SDKPackageInstallation
Pythonhmdlpip install hmdl
JavaScript/TypeScripthmdlnpm install hmdl

Next Steps