Observed arrival · 2026-10-21
V3R1T4S: AI agents take on open problems
A public project invites people to point local AI agents at open problems in mathematics and science, then submit results with checkers for others to run.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- AI-agent builders and open-problem hunters
- Worth noticing
- Submissions need a runnable checker, user approval, and maintainer re-check; the project says results are not peer reviewed.
Field notes
The workflow runs through a local agent rather than a plain chat app: the page names Claude Code, Codex, and Gemini CLI as examples, and says the agent asks six setup questions. The homepage says runs often produce no result; a submission needs a checker, user approval, and maintainer review before appearing on the board. Its weekly update reports 280 entries awaiting review.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
◉HumanPersonal, local, civic, or handmade
One card from the complete issue