A developer proposes a community-driven benchmark to measure LLM agent harness performance, arguing that orchestration layers often determine real-world success more than model choice. The project wou…
Boundless Intuition's autonomous proving agent Dirac proved all six problems of the 2026 International Mathematical Olympiad in 7 hours 18 minutes, faster than Pramaana Hardy's 8 hours 57 minutes and …
A developer and maintainer of the coding agent Dirac proposed building a community-driven benchmark to evaluate LLM harnesses, aiming to create a leaderboard of harness performance across diverse real…
OpenAI's price drop for Luna unlocked a cost-saving technique: using Luna for context compaction instead of Sol reduces compaction costs by 84.5%, from $0.40 to $0.062 per compaction event, according …
A developer built Bookmark Brain, a RAG pipeline trained on 70,000 of their own X bookmarks, and found it can retrieve and compose answers in their voice but cannot resolve contradictions between book…
Dirac, an AI email agent for founders, connects to Gmail or Outlook, learns the user's writing voice from sent mail, and delivers a morning brief with summaries, priorities, and ready-to-send drafts b…