Non-coder builds multi-agent system that found rule loopholes on day one
akanten · reddit · 2026-08-23
A non-technical operator from Norway built a public site where AI agents from different vendors (currently Claude and ChatGPT) check each other's work. An agent picks a skill, applies it to a real task, and files a GitHub PR. A skill is only promoted after two agents on different underlying models validate it independently.
Results from the first day include:
- Claude and ChatGPT independently flagged a real hole in the governance rules the author had written hours earlier, leading to an immediate fix.
- A friend's independent agent (Hermes) returned 23 concrete findings, most of which were fixed in the same session.
- The models did not just agree to be nice; they found each other's mistakes, and the history remains public.
Related event: Multi-Agent Cross-Checking System Detects AI Hallucinations(2 posts)→
More from coding & agent
- YC F26's Praxis launches AI agents to run law firm ops end to end — ycombinator · 2026-10-16
- Developer builds inter-agent messaging for his AI agents — complete with message undo — baditaflorin · 2026-10-16
- iOS 'dots' workflow: one Siri command turns screenshots into a background shopping and saving agent — nickbaumann_ · 2026-10-16
- Your widget tree is already a list of options: where Jev fits in Flutter — rseroter · 2026-10-16
- How OpenAI's Codex went from failed side project to 40M weekly users in under a year — danshipper · 2026-10-16
- Photon 2.0 launches unified messaging API for AI agents across iMessage and more — Scobleizer · 2026-10-16