Developers can now debug and evaluate AI agents locally with Raindrop's open source tool Workshop

Raindrop AI launched Workshop, an MIT-licensed local debugger and evaluation tool for AI agents.

Workshop records agent traces to a local .db file and dashboard, giving developers private, on-device observability.

It supports multiple languages and agent frameworks, including Claude Code, Cursor, Devin, and OpenCode.

A self-healing eval loop can detect failures, rerun tests, and help agents fix behavior until checks pass.

The tool aims to speed iteration and make autonomous AI workflows easier to trust and maintain.