Policy dependency / Stack layer
A replay of 68,266 real Claude Code requests says plain LRU beats the clever KV-cache policies
GitHub
Policy dependency / Stack layer
Microsoft publishes a 37-page 'humanist AI code of conduct' for its MAI models and opens it to public consultation
The Verge (corroborated by Unite.AI and TechBriefly)
Stack layer / Threat pattern
Pattern: Anthropic's own bar is that Claude-written production code gets reviewed harder than human-written code
Simon Willison
Policy dependency / Stack layer
git-ai is a Git extension that tracks which code in your repo was AI-generated, at 140 open PRs to 71 issues
GitHub Trending
Stack layer / Update thread
Open Letter to Amodei Asks for One Law Instead of Evaluators: Any Model Sold to the Public Must Ship Open Weights
Jake Gold / Hacker News (290pts, 96 comments)
Stack layer / Threat pattern
Snyk put its agent-skill scanner behind a free web page called Skill Inspector
Snyk Labs
Policy dependency / Contrast
Former EPA officials say data center buildout is being exempted from pollution rules
The Verge
Policy dependency / Stack layer
China Drafts Embodied-AI Data Standards After 70+ Robot Training Grounds Went Up Without Them
The Next Web