Malicious Claude Code extension with backdoor found in Cursor marketplace flags AI security wild west
dbasch · x · 2026-10-31
dbasch warns of a new "wild west" for hackers: a malicious Claude Code extension with a backdoor was injected into Cursor's marketplace, giving attackers instant access to every developer's machine who installed it.
Key points:
- Anthropic (2021) and Cursor (2022) are used everywhere but are too young to have mature security practices — it took Apple and Microsoft decades, and you can't speedrun security.
- Organizations now run a grab bag of AI tools (Claude, ChatGPT, Codex, etc.), generating code at insane speed on top of tools maintained by people doing the same.
- His advice: inventory every AI tool your org uses, review its access levels, and prepare a rapid-response plan for compromise — no existing security playbook covers this yet.
More from coding & agent
- Meta's coding model Muse Ember goes live on OpenRouter for agent sub-task fan-out — alexandr_wang · 2026-11-10
- Meta's Muse Ember coding model is now available in OpenCode — alexandr_wang · 2026-11-10
- OpenRSI scales recursive self-improvement to SGLang-Omni, +7.6%–344% throughput via 14 merged PRs — ChengleiSi · 2026-11-10
- OpenCode cuts memory 41% and OpenClaw starts 3.1x faster with Bun AoT — steipete · 2026-11-10
- Vibe-coded ambient music DAW 'Ambient Town' in a week with Push support — kieranklaassen · 2026-11-10
- GR00T 2's WAM API lets reasoning agents steer robots by pointing in the camera view — yukez · 2026-11-10