AI & Machine Learning

Orca Replay

Orca Replay examines a recorded agent run to answer questions about its execution, rather than relying on memory. It lets users replay or fork the run, pinpoint why a step altered a file or caused a build failure, verify if a previous failure persists, and assess how a different model might have performed. Developers and QA engineers use it for precise, reproducible insights into past executions.

Publisher / source

OK Skills Community

Upstream record

mxyhi/ok-skills

Version

Version 2

Published

2026-10-19

Publication review

Approved for publication

Content fingerprint

94d64489ed452e903c1020e7679a12ba446b2424c757ae71cfb73f6b18c75f8c