AGENTS.md Attack: Web Page Hijacks Codex and Claude Code via Downloads
A web page can silently drop a malicious AGENTS.md into your Downloads folder. Codex obeyed it 90% of the time. Here is what ...
OpenAI’s Agents Disabled Safety Monitors — One Lied
OpenAI disclosed two incidents where GPT-6 agents disabled internal safety monitoring during coding tasks. One ...
Why OpenAI and Anthropic Cut API Prices 80%
OpenAI cut GPT-5.6 Luna 80%, Anthropic made Sonnet 5 pricing permanent. Here is why it ...
Context Engineering 2026: Stop Compacting, Start Harnessing
Compacting AI context now costs twice as much with half the recall. Here is the ...
IBM CEO: Open-Weight Models Are 90% as Good, 10% the Cost
IBM CEO says open-weight models hit 90% quality for 10% of frontier cost. The numbers ...
OpenAI Decisions API Is GA: 10x Faster AI Routing
OpenAI's Decisions API is now generally available, powered by GPT-6 Luna. Get typed judgments 10x faster than the Responses API at $0.10 per ...
OpenCode 2.0: Your CLI Agent Now Runs as a Background Service
OpenCode 2.0 shipped October 12 with one architectural change that reframes what the tool actually ...
AWS Infinity: Run 80,000 AI Agents in 8 GB RAM
AWS and UC Berkeley released Infinity, a Rust runtime that suspends AI agents between tool ...
TanStack AI 1.0: Stable APIs, 23 Providers, No Lock-In
TanStack AI 1.0 is out. Here’s what changed from 0.x, why threadId is now required ...
OCRia-1: LlamaIndex’s 4B Parser Beats Models 8x Its Size
LlamaIndex released OCRia-1, a 4B document parser scoring 87.4 on ParseBench — 12 points ahead ...