arielshad/balagan-agent

Chaos Engineering for AI Agents

42
/ 100
Emerging

This helps AI engineers and SREs rigorously test their AI agents before deployment. It takes your existing AI agent, simulates real-world problems like tool failures, delays, or corrupted data, and then shows you how resilient your agent is. The output includes metrics like recovery time and a reliability score, allowing you to identify and fix vulnerabilities proactively.

Available on PyPI.

Use this if you are building AI agents for production and need to ensure they are robust and reliable when external services or data sources behave unexpectedly.

Not ideal if you are only experimenting with AI agents in non-critical environments where reliability and fault tolerance are not primary concerns.

AI-reliability agent-testing chaos-engineering SRE-practices production-AI-validation
Maintenance 10 / 25
Adoption 4 / 25
Maturity 20 / 25
Community 8 / 25

How are scores calculated?

Stars

8

Forks

1

Language

Python

License

Apache-2.0

Last pushed

Feb 15, 2026

Commits (30d)

0

Dependencies

2

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/arielshad/balagan-agent"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.