Three coding-agent security patches ship this week: Mistral Vibe, Jan, DeepSeek Harness
alvelda · x · 2026-11-10
AgentFieldKit's weekly roundup flags three coding-agent security fixes worth installing immediately:
- Mistral Vibe 3.0.3: a crafted repo could run a program before the trust prompt appears
- Jan 0.8.8: a dangling-symlink write allowed escaping the workspace
- DeepSeek Harness 0.1.2-alpha.2: an unauthenticated local RPC endpoint
The advice: update now, and open untrusted repos with care — a common exposure across agent tooling.
More from coding & agent
- Developer dedicates a Mac mini as a home for his AI agents — ow · 2026-11-10
- Hyo adds Agent Dock to Obsidian: pin and monitor all your AI chats in one sidebar — evielync · 2026-11-10
- GPT 6.1 agent designs a cascadable broadband half-adder in a 23-minute run — CatAstro_Piyush · 2026-11-10
- Claude Code's PM #2 shares how to do devtools PM in the AI era — EricBuess · 2026-11-10
- Dev demos AI that edits documents live while you think out loud — jungofthewon · 2026-11-10
- Revolut cut its AI coding bill by 90%: 256K context enough, cache 90% of tokens — alvelda · 2026-11-10