News for hackers. Stuff that matters.
Re:Show HN: R3 – Review AI agent artifacts locally by rickye26 on Tuesday November 17, 2026 @05:20PM (#50539069)
Hi HN, author here. I built r3 to solve pain points I felt when using AI coding agents. I frequently let agents to write summary report or design doc in html, however the chat interface is not a good match for giving feedback for long doc. I found that I often need to paste text from the html to reference it in the chat box, and it's so painful to track status of feedback on 10+ different places of the doc.

r3 is like Claude Artifacts, which lets you to directly comment on the web page, and send these feedback back to the agent by clicking a button. r3 runs locally and works with every agent interface/harness.

Have a try and I'd love to hear your thought on it!

There are no comments on this comment yet.