Running in production · Independent · 19 Oct 2026

latitude-dev/latitude-llm

Open-source observability for AI agents. Find where your agents fail, dispatch your coding agent to fix it, and verify the fix against real traces.

← Running in production

In simple English

When you build software powered by artificial intelligence, your automated helpers often make mistakes that are hard to catch. This tool watches your app, records every step the helper takes, and points out where it broke. If something fails, it can send a message to a coding assistant to write a fix for you. You can use their free cloud service, or run the whole thing on your own server. It costs nothing to try, with limits on monthly data storage.

What it is

Latitude is an open-source observability platform for artificial intelligence agents that tracks execution traces, groups failures, and dispatches coding assistants to fix them. It captures multi-turn sessions and tool calls, then verifies fixes against real regression test datasets.

What it is good for

Solo builders who ship software driven by large language models and spend too much time hunting down why an agent went off the rails. It takes the pain out of debugging complex multi-step prompt chains.

Stars4.7k
LicenceMIT
Latestcli-4.0.1 (2026-10-09)checked 19 Oct 2026
Last code change2026-10-19
Written inTypeScript
Setup effort
3 / 5
Payoff
4 / 5
Good
  • Captures complete execution paths, multi-turn sessions, and tool calls with a single line of telemetry.
  • Automatically groups failing traces so you can see what breaks most often and why.
  • Dispatches coding agents like Claude Code or Cursor with full context to write patches and open pull requests.
Watch for
  • Self-hosting a full production cluster requires managing Docker containers and backend infrastructure.
  • Heavy reliance on external artificial intelligence providers and third-party coding agents for the self-healing loop.
  • Node.js (a system that lets you run JavaScript outside a web browser) environment required for telemetry installation and TypeScript configuration.
Replaces

manual log reading and trial-and-error prompt debugging

Works well with

OpenAI API

Similar repositories
  • langfuse/langfuse36k stars

    An open-source language model engineering platform focused on tracing, prompt management, and analytics.

  • promptfoo/promptfoo26k stars

    A tool for testing and evaluating prompts and artificial intelligence applications against regressions.

  • Arize-ai/phoenix12k stars

    An open-source observability tool for evaluation, troubleshooting, and fine-tuning generative applications.

Also reviewed on Repo Radar
  • nicolargo/glances34k stars

    A live view of one computer's health, in a terminal, in a browser, or through an AI assistant.

  • dokku/dokku32k stars

    A small platform you install on one server, which builds and runs an app every time you push code to it.

  • netbox-community/netbox22k stars

    A database of what a network and its hardware are supposed to look like, which other programs can read.

Install
npm install @latitude-data/telemetry

Independent: reviewed for Repo Radar outside the daily editions. Every independent repository

On Stack Radar: AI & Automation › AI Development & Evaluation › AI Observability.

Get the next edition in your inbox

A dozen repositories, opened and checked. The licence read, the last release dated, and the ones that did not make it named with the reason. It is the half most lists leave out.

No tracking pixels. One click to leave. The archive stays free either way.

We use your address to send the edition and nothing else. Confirm by email, leave in one click. How we handle it.