ARC-AGI-N: open-source tool surfaces unsolved science problems and preps context for your AI agent
Physical_Pepper6294 · reddit · 2026-09-08
Inspired by Claude improving the zeta-zero lower bound from 41.6% to 67.2% while attempting the Riemann hypothesis, an 11-day Lean formalization of Wiles's proof, and GPT-Astra scoring 99.9% on ARC-AGI-3, a developer built ARC-AGI-N, an open-source research tool.
What it does:
- Searches papers and the web for open problems in math and science, showing sources in real time
- Each problem includes the question, background, source material and a suggested starting point
- Copy a full-context prompt into your agent, or run DeepResearch first to generate history, prior attempts and a 72-hour work plan
- Browse problems on an interactive Mapbox globe, plus a log of AI-assisted discoveries and formalizations
Built with Next.js/React/TypeScript, OpenAI Luna and Valyu's search/DeepResearch APIs. Open-source, self-hostable with your own keys, with a hosted version available.
The author stresses it won't turn prompts into proofs—the goal is making it easy to find a worthwhile problem and give agents enough context to start.
Related event: ARC-AGI-N: Open Tool Auto-Discovers Open Problems for AI Agents(2 posts)→
More from coding & agent
- ReCache cuts agentic KV memory 92.43% and speeds first token 3.655x with minimal accuracy loss — techNmak · 2026-09-08
- ReCache reuses tool-schema KV states, cutting agent memory 92% with little accuracy loss — techNmak · 2026-09-08
- Weaviate's year of agent memory: why store-everything RAG breaks three ways — victorialslocum · 2026-09-08
- A simple Next.js app built 2 months ago already has 48 vulnerabilities — haydendevs · 2026-09-08
- GPT-6 Astra 'feels like AGI': $1,000 hands-on across real client work, gaming and computer control — PrajwalTomar_ · 2026-09-08
- Google open-sources Artemis, an Android-automation agent scoring 99%+ on AndroidWorld — Shruti_0810 · 2026-09-08