Skip to main content

When your agent fails, this tells you exactly why.

Project description

AgentAutopsy

When your agent fails, this tells you exactly why.

Why this exists

Every time an AI agent fails, you get a useless stack trace. No context. No reason. No fix. AgentAutopsy gives you the exact failure step, root cause, and a verified fix — automatically.

demo

Python License Zero Config Works with

CLI

agentautopsy runs # see all agent runs agentautopsy replay # replay any failure agentautopsy stats # fix cache stats

Install

pip install agentautopsy

Examples

# Basic usage
import agentautopsy
agentautopsy.watch()

# LangChain
from agentautopsy import get_callback_handler
handler = get_callback_handler()
agent.run(input, config={"callbacks": [handler]})
# Slack alerts
export AGENTAUTOPSY_SLACK_WEBHOOK=https://hooks.slack.com/...

# Web UI
agentautopsy ui

# CLI
agentautopsy runs
agentautopsy replay <run_id>
agentautopsy stats

Usage

import agentautopsy
agentautopsy.watch()
# your existing agent code here — nothing else changes

AgentAutopsy automatically intercepts every LLM call, detects failures, finds root cause, outputs a verified fix, and caches it for next time.

Why AgentAutopsy vs LangSmith / Helicone?

Feature AgentAutopsy LangSmith Helicone
Works offline
Zero config
Replay failed runs partial
AI debug assistant
Prompt diffing partial
Divergence detection
Free and open source partial
No cloud required

Setup

Windows: set ANTHROPIC_API_KEY=your-key-here Mac/Linux: export ANTHROPIC_API_KEY=your-key-here Get your free key at console.anthropic.com

Set AGENTAUTOPSY_SLACK_WEBHOOK=your-webhook-url and AgentAutopsy will automatically alert your Slack channel when any agent fails.

Quick start

Create test_agent.py and paste this:

import agentautopsy
agentautopsy.watch()

Run: python test_agent.py

Works with

OpenAI, Anthropic, LangChain, any framework using openai or anthropic

Requirements

Python 3.11+, ANTHROPIC_API_KEY

License

Apache 2.0

Roadmap

  • VS Code extension
  • GitHub Actions integration
  • Multi-agent tracing
  • Auto-fix applier
  • LangChain support
  • Slack alerts
  • Web UI
  • Prompt diffing
  • Divergence detection

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

agentautopsy-1.4.0.tar.gz (1.4 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

agentautopsy-1.4.0-py3-none-any.whl (43.2 kB view details)

Uploaded Python 3

File details

Details for the file agentautopsy-1.4.0.tar.gz.

File metadata

  • Download URL: agentautopsy-1.4.0.tar.gz
  • Upload date:
  • Size: 1.4 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.5

File hashes

Hashes for agentautopsy-1.4.0.tar.gz
Algorithm Hash digest
SHA256 2a4d8a278772dccc04c29007785643d307e74c8117b94c74c7f1ac93c5fd3eb7
MD5 313132690ec618a069ee818132d6c81d
BLAKE2b-256 45ada741fa0800d4e8df23f6c6866df30adbb066093ba72bc5f7c202d3105f6c

See more details on using hashes here.

File details

Details for the file agentautopsy-1.4.0-py3-none-any.whl.

File metadata

  • Download URL: agentautopsy-1.4.0-py3-none-any.whl
  • Upload date:
  • Size: 43.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.5

File hashes

Hashes for agentautopsy-1.4.0-py3-none-any.whl
Algorithm Hash digest
SHA256 9f0730e8b6ac2f52c45b729f1536f7f24c30b07653cad085134a555041756258
MD5 a8ab8c57a4c3d38054c0424a72d89556
BLAKE2b-256 a1b525db52e2828f7ae86eb6440d3da65fddd00a734c4b6f8e2593e8c5b21c2e

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page