Moonshot AI's Kimi Code CLI hits GitHub trending after subagent
Kimi Code CLI 2.3.2 stops automatically spawning subagents, shifting how terminal-based AI coding agents handle multi-file tasks.
The top 0.01% of AI knowledge, and what it changes for your work
Daily AI and technology coverage — models, chips, agents and research, and what each development actually means for you and your business.
732 stories
Kimi Code CLI 2.3.2 stops automatically spawning subagents, shifting how terminal-based AI coding agents handle multi-file tasks.
GenOffice, a free open-source AI Office suite trending on GitHub, lets an AI agent edit real .docx and .xlsx files on your own machine.
An October 23 preprint claims NgramRL commits up to 7.77 tokens per forward in warm-cache tests, beating prompt-lookup 7×, no auxiliary model.
An open-source scraper for Xiaohongshu, Douyin and five other Chinese apps is trending on GitHub, but its own disclaimer clashes with its paid Pro
A deep learning surrogate approximates BurnP3+ wildfire simulations 806x faster on CPU, cutting Canada's hazard mapping from weeks to seconds.
AllenAI's OlmoEarth v1.3 releases precomputed satellite embeddings for the entire planet at 10-meter resolution, with fully open weights.
Herdr, a Rust terminal runtime trending on GitHub, keeps Claude Code and Codex sessions alive when SSH drops, with 45,735 stars.
The Agent Skills specification gives AI agents new capabilities through folders and markdown files, gaining developer traction on GitHub.
BerriAI's LiteLLM AI gateway reached 61,877 GitHub stars, routing 100+ LLM APIs through a Rust core, but its 8ms latency claim remains unverified.
Fleet AI researchers built Fig to systematically test whether AI models will deceive, betray or sabotage in multi-agent game environments.
Carnegie Mellon's AgentPhishing benchmark shows AI email agents that catch standard phishing still fall for attacks tailored to their behaviour.
A TU Berlin preprint rewrites attention as damped Hamiltonian dynamics, lowering validation loss on small models but reporting no numbers.