Rill
Browser evidence for you and your AI agents
Start new thread

Rill - Browser evidence for you and your AI agents

by•
Your agent says it works. See for yourself. Rill records your AI agent’s browser run with video, console logs, network requests, and an interaction timeline. Replay the flow, investigate failures, and share the whole run in one link. Use it to review a fix before shipping, attach evidence to a pull request, or give the next person or agent the context to continue.

Add a comment

Replies

Best
Maker
📌
Hey Product Hunt! We built Rill around a simple question: when an AI coding agent says a browser flow works, how do you check what actually happened? Rill gives you a recording of the run alongside console logs, network requests, and an interaction timeline, all in one shareable link. You can watch the flow, investigate an error, and pass the evidence to a teammate or another agent. Set up the CLI, give your agent the recording instructions, and ask it to record a browser flow. The agent runs the flow and returns a link you can review. We’d love to hear how you verify agent-built features today. What evidence would make you more confident approving a change?
OpenAI
GPT-6 Astra Challenge
What became possible in your product with Astra that was not practical before?
Maker
Astra helped us take Rill from coding through launch preparation in one workflow. We used it to help code the product, set up Blender to create product launch videos, and drive edits in CleanShot Studio Mode. This made it more practical for us to build the browser-evidence tool and produce the videos needed to demonstrate it. Rill captures an AI agent’s browser run with video, console logs, network requests, and an interaction timeline, so people can inspect what actually happened. Our use of Astra was in building and presenting Rill; the product gives people and agents evidence they can review and share.