Introduction

Neatlogs is an observability platform for AI agents. Instrument once, then inspect every trace your agent produces: LLM calls, tool invocations, token counts, latency, and more in a shared dashboard your whole team can use.

This is all it takes

Add a few lines to your existing code:

import os
import neatlogs

neatlogs.init(
    api_key=os.environ["NEATLOGS_API_KEY"],
    endpoint=os.environ["NEATLOGS_ENDPOINT"],
    workflow_name="my-agent",
    instrumentations=["openai"],
)

That's it. Every call your agent makes is now captured as a trace and visible in your dashboard. No changes to the rest of your code.

For a complete walkthrough, see Your First Trace.

Why Neatlogs

Agent failures don't throw exceptions. They produce wrong outputs, miss tool calls, or hallucinate. Diagnosing them requires seeing what the model was given, what it decided, and what each step returned. That context lives in traces.

Neatlogs is built around traces as the primary debugging artifact. Every run is fully captured. Engineers can inspect raw span data; non-engineers can search, comment, and flag issues without needing to understand the underlying data model.