Documentation

AgentIdem documentation

Learn how to test state-changing AI agents under retries, duplicate delivery, lost acknowledgements, and other controlled fault conditions.

Documentation

Choose where to start.

SDK reference

Integrate AgentIdem directly into Python applications with reads, writes, traces, fault scenarios, invariants, and test execution.

Command-line reference

Run AgentIdem from the terminal or CI using test, record, replay, and version commands.

Quick links

These pages cover the main pieces you need to start using AgentIdem without reading the full reference in order.

Install AgentIdem from PyPI, then start with the AgentIdem Nebutex API quick start.