Developers can now debug and test AI agents in the environment with Raindrop’s open source tool Workshop
Realization startup Raindrop AI’s new open source, MIT License "Workplace" tool, introduced today, gives developers something they may have been looking for, perhaps unconsciously, since the AI era began in earnest last year: a local debugging and testing tool designed specifically for AI agents, allowing devs to see all traces of what their agent was doing in a lightweight, structured Query (dSQL) database file.
It acts as a local daemon and UI that broadcasts all tokens, tool calls, and decisions to a local dashboard—usually hosted by localhost:5899– when it happened. By visiting their local host, developers can see everything their agent has been up to – including bugs or errors – and pinpoint what went wrong, when, and ideally, why. Everything is stored in a single .db file, which takes up relatively little memory, according to a direct X message VentureBeat received from Ben Hylak, Raindrop’s founder and CTO (and former Apple and SpaceX engineer).
This real-time telemetry eliminates common polling delays and addresses developers’ growing concerns about the privacy of sending location traces to external servers.
The tool is available for macOS, Linux, and Windows. It can be installed with a one-line shell command that automates the binary placement and PATH configuration of bash, zsh, and fish shells. For developers who choose to build from source, the repository is hosted on GitHub and uses the Bun runtime.
Product: establishing a self-sustaining eval loop
The prominent feature of the platform is "self-contained eval loop," which allows coding agents like Claude Code to read traces, write evals against the codebase, and fix broken code automatically.
In a practical application, if a veterinary assistant agent fails to ask the necessary follow-up questions, Workshop takes the full lead. Claude Code then reads this trace, writes some eval, identifies a logic error in the notification or code, and reruns the agent until all assertions pass.
Compatibility and ecosystem integration
Workshop is compatible with a wide range of programming languages, including TypeScript, Python, Rust, and Go.
It also includes popular SDKs and frameworks such as Vercel AI SDK, OpenAI, Anthropic, LangChain, LlamaIndex, and CrewAI. It is also designed to work seamlessly with various coding agents, including Claude Code, Cursor, Devin, and OpenCode.
Licensing and public implications
The workshop is released under the MIT License, ensuring that it remains free and open source for all users. This permissive licensing is intended to encourage public contribution and allow business users to maintain data sovereignty.
Hylak noted in X that the tool was designed to provide a "you are fine" how to fix agents in place, changing the way their team and first customers build independent systems.
To celebrate the launch, Raindrop offered limited edition physical merchandise to users who installed the tool and used it. "dripping" commandment.
