Stories

  1. Image: BleepingComputer

    Cursor says key stolen via Fireworks AI breach led to malicious Claude Code extension in its marketplace

    Cursor has confirmed a security breach that allowed attackers to distribute a malicious version of Anthropic's Claude Code extension through Cursor's extension marketplace. The company said an attacker obtained an access key Cursor had issued to AI infrastructure provider Fireworks AI, which was exposed in a breach targeting Fireworks. The attacker used the key to copy data from a read-only resource, Cursor said in a Friday update.

  2. Image: 9to5Mac

    TVPort brings Nintendo Switch games to Apple TV using Claude Code

    A Redditor has released TVPort, an open-source project that runs Nintendo Switch games natively on Apple TV, using Claude Code in the porting effort. Super Mario Odyssey and Animal Crossing: New Horizons are running, while work continues on The Legend of Zelda: Tears of the Kingdom. The project is presented as another example of Claude Code being used to bring software to different platforms.

  3. Image: Ars Technica

    Anthropic updates Claude Code instructions to block recreations of existing video games

    Anthropic has updated Claude Code’s system prompt to discourage recreating specific existing video games, while still allowing users to build original games and games in a genre. The rule says not to make a recreation even if it is renamed or described as original. XDA independently verified the instruction in Claude Code version 2.1.314, released October 30. The change is unlikely to stop the broader wave of AI-assisted game porting and copying, one account argues.

  4. Image: Tom's Hardware

    Anthropic restricts Claude from recreating and reverse-engineering existing video games

    Anthropic has restricted Claude from recreating specific existing games, with users reporting refusals on game-cloning and reverse-engineering tasks. A rule shared by a user says Claude can build original games and genre-based projects, but should not recreate an existing game even under a different name. Developers also report the restriction blocking legitimate work, including troubleshooting and recreating their own earlier work. Anthropic has not officially confirmed the quoted rule text.

  5. Image: Ars Technica

    Tampered Claude Code extension on Cursor marketplace planted persistent SSH backdoors

    A malicious actor used access to Cursor’s extension marketplace to distribute a tampered Claude Code extension, Anthropic said. The package, version 2.1.313, was available for about 11 hours on October 29, and security reporting says it could expose up to 5,000 developers to malware that installed persistent SSH backdoors. Anthropic said its systems and employee laptops were not compromised; its official distribution channels were unaffected.

  6. Image: GitHub Changelog

    GitHub launches Copilot Fusion for automated model selection and orchestration

    GitHub announced Copilot Fusion at its GitHub Universe event in San Francisco, introducing a system that selects and orchestrates models for coding tasks. The company says Fusion Max matched Claude Code with Opus 5.5 Max on benchmarks while costing about half as much in average API fees and taking about 40% less time per task. The system is generally available, allowing developers to delegate model choices while retaining responsibility for their work’s direction and outcome.

  7. Image: Phoronix

    Linus Torvalds removes legacy __get_user() and __put_user() functions from Linux 7.4

    Linus Torvalds has merged a 48-part patch series removing the kernel’s __get_user() and __put_user() functions, with the changes destined for Linux 7.4. The primitives were historically used to access user-space memory from the kernel, but the older approach has been superseded. The cleanup addresses code that has remained in the large kernel codebase despite that change.

  8. Image: The Register

    ARTEX AI Pentesting Tool Used in Data-Theft Campaign Against South Korean Financial Firms

    CrowdStrike Intelligence identified a campaign targeting South Korean financial organizations that used the AI penetration-testing tool ARTEX. The activity ran from late September to early October 2026 and resulted in data exfiltration from firms in the country. Coverage describes the suspected attacker as Chinese-speaking and financially motivated, and reports use of large language models alongside ARTEX.

Topics

Publishers