Claude Telemetry

Observability for Claude Code

claude_telemetry wraps the Claude Code CLI with full observability. Every tool call,
token count, and cost flows to your OTEL backend as structured traces.

The Problem

Claude Code works beautifully in your terminal. But when you run agents headless—in
CI/CD, cron jobs, production, or remote servers—you lose visibility into what's
happening. You can't see which tools were called, don't know token usage or costs, have
no context when debugging failures, and can't optimize without re-running locally.

Headless environments lack the rich console output you get during development.

The Solution

claude_telemetry is a thin wrapper around the Claude Code CLI that adds observability.
It passes through all Claude Code flags unchanged while capturing execution traces.
Works with any OTEL backend—Logfire, Datadog, Honeycomb, Grafana. See exactly what
happened in production, debug remote failures without local reproduction, and track
costs to optimize expensive workflows.

Your agents become observable whether local or headless.

Quick Start

The simplest way to add observability: swap claude for claudia on the command line.

That's it. Every flag you use with claude code works with claudia. The behavior is
identical, but now you get full traces in your observability platform.

For Developers

If you're already using claude code in your workflow, switching to claudia gives you
instant observability with zero behavior changes.

Command-line usage:

In scripts and automation:

The result:

When you use claudia, you get the exact same output in your terminal as claude code.
But now your observability platform shows every tool call, token count, cost, and
timing. Debug headless failures, track production costs, and optimize expensive
workflows—all…