The Signal runs on the AI-Generative Platform.See the platform →

AI-GENERATIVE.ORG

AI-Generative Signal

Watch your Signal brief

Duration: 1:43

The Signal Diagram showing a secret-bearing command saved in Claude Code local settings, then exposed only if the file is committed and published to a public repository. A machine-local Git exclusion does not travel with the repository.

Claude Code Settings Files Expose Secrets on GitHub, Researchers Report

Claude Code’s persistent approvals can preserve embedded credentials in settings files that later enter public repositories, Apiiro reports. The findings distinguish local-file exposure from a broader investigation, while Anthropic’s discussed safeguards remain proposals rather than a verified shipped fix.

 Read the full story →
The Pulse

The backlash about data centers might really just be about electricity

To get stories like this in your inbox first, sign up here.

MIT Tech Review AI · 2026-11-09T19:35:33.000Z  Read the story →

Apple’s $50 Intelligence Usage Packs extend access to some AI features

After publishing a support article in early September outlining usage limits for some Apple Intelligence features, the company has updated Logic Pro with an option to "Buy More Usage" for a one-time purchase price of $49.99, as spotted by 9to5Mac. It has a built-in AI assistant to do things like "start a new song, set up a vocal track from a single prompt, or refine an active project."

The Verge AI · 2026-11-09T19:09:58.000Z  Read the story →

Anthropic launches new Claude Fable and Mythos 5.2 models

Anthropic is announcing two new models today: Claude Fable 5.2 and Claude Mythos 5.2, which it says are designed to take on "daunting" technical challenges. The company claims the new models show improvements in coding, reasoning, and vision, and can generate outputs faster than the 5.1 versions of the Fable and Mythos models.

The Verge AI · 2026-11-09T19:00:00.000Z  Read the story →

The Feature
Today's Feature

How to compare agent tuning with search-tree design

Duration: 2:55

Flex-RSI evaluates changes to an agent's working setup, while TreeFlex describes navigation through saved solution paths. This comparison explains their different mechanisms, keeps software-engineering results in context, and turns the distinction into a practical specification for evaluating 3D agents.

 Watch on the site →
The Move

Read the primary coverage behind "The backlash about data centers might really just be about electricity" (MIT Tech Review AI) directly, rather than relying on this summary alone -- useful for anyone tracking policy_security closely.  Read: MIT Tech Review AI →

Reuses this edition's own freshly link-verified Pulse evidence rather than introducing a new, unverified link.

The Watch
  • Whether "Apple’s $50 Intelligence Usage Packs extend access to some AI features" (The Verge AI) develops into a broader pattern worth tracking, or remains an isolated development.  Follow: The Verge AI →

Want more of these, every day?

Sign up for AI-Generative Signal and get the daily briefing straight to your inbox.

Sign up for the Signal →

This content was assisted by AI tools; editorial responsibility rests with AI-Generative.

Canonical: https://www.ai-generative.org/news/claude-code-settings-files-expose-secrets-on-github-researchers-report

Edition canonical: https://www.ai-generative.org/signal/2026-11-09 · Browse the Signal archive · The Move archive