AI & LLMs
Claude Opus 4.7 agents invent games on a virtual island
An MIT study left 13 Claude Opus 4.7 agents alone on a virtual island for 30 hours. They invented games, then beat fresh copies on new tasks.
Claude Code news and analysis
Claude Code releases, defaults and behavior changes, covered for the developers who run it every day: what a new default does to your commits, what it costs, and how to opt out.
Coding
NIST P-256 curve seeds traced to a 1997 lrand48 run
Anthropic traced the NIST P-256 curve seed to an lrand48 run started August 16, 1997, claiming a $36,864 bounty. The finding is not a break of the curves.
Coding
Torvalds uses Claude Code to rework the Linux page cache
Linus Torvalds posted eight Linux kernel patches made with Claude Code. On his laptop, 8-thread small reads used about 4x fewer CPU cycles each.
Dev Stack
YSERVER 1.7 adds Vulkan 1.2 and cuts 4K VRAM use
YSERVER 1.7 cuts 4K VRAM use from 2-4GB to about 1.2GB, near X.Org's 0.9GB, and adds Vulkan 1.2 support for older Intel graphics.
Dev Stack
Claude Code 2.1.287 adds Claude Mods for deeper plugins
Claude Code 2.1.287, out October 1, lets plugins change the agent's deeper behavior through Claude Mods and adds an opt-in side agent that flags mistakes.
AI & LLMs
Claude for Government reaches GA under FedRAMP High
Anthropic's Claude for Government is generally available to U.S. federal and state agencies under FedRAMP High, with no seat fees and hard spending caps.
Dev Stack
Claude Code AGENTS.md support needs telemetry turned on
Claude Code 2.1.277's AGENTS.md support sits behind a remote feature flag, so it silently stops working when telemetry is off, a GitHub issue shows.
Dev Stack
Claude Code, Codex, and Cursor agree on a tool only 42% of the time
A study of 16,893 coding-agent sessions found Claude Code, Codex, and Cursor pick the same third-party tool only 42% of the time. Stripe beat PayPal in every session where both qualified.
AI & LLMs
Claude Fable 5.1 cuts cache reads 75% and keeps token prices flat
Anthropic's Fable 5.1 holds input and output prices at $10 and $50 per million tokens, and drops cache reads from $1.00 to $0.25. Mythos 5.1 stays invitation-only.
AI & LLMs
Anthropic's new standard lets AI agents run lab equipment
Anthropic opened a research preview of the Model Hardware Standard on August 27, 2026. It gives agents one way to drive microscopes, robot arms and lab tools.
Dev Stack
Claude Code puts a session link in commits unless you opt out
Claude Code appends a Claude-Session: trailer linking to its own chat to commits and pull requests. It defaults to on, and developers want it opt-in.
Coding
JetBrains: developers say agents now write 47% of their code
A JetBrains survey of 15,509 developers finds agents fully write 47% of code on average. 90% use an agent weekly, and Claude Code leads adoption at 39%.
The weekly digest
One email every Friday with the week's top stories from all six desks: AI, dev tools, coding, the tech industry, startups and what's next. Free, no spam.