After the Hackening, he set Claude Code to update everything — and keeps getting approval prompts
TheZvi · x · 2026-10-22
AI researcher Zvi shares that after the recent Hackening security incident, he has Claude Code running updates on everything on his computer — and a day later he's still being randomly interrupted for periodic approvals. A relatable snapshot of agent autonomy vs. human-in-the-loop friction.
More from coding & agent
- Recruiting tool binary.so launches MCP and Claude Connector, used by 600+ startups — Composer-Chemical · 2026-10-22
- Claude desktop can clean your PC: 7-step computer use prompt checklist goes viral — aitrendz_xyz · 2026-10-22
- Building breakable cities with Opus 5.5 — open-source and playable now — smith2008 · 2026-10-22
- 7-step recipe turns Claude Code into a second brain with Gmail, Drive and a knowledge graph — anirbanbandyo · 2026-10-22
- Three bots run a self-writing wiki in one group chat, growing a second brain from use — anirbanbandyo · 2026-10-22
- skyos.ink: an MCP memory server that keeps project facts as dated notes, not one summary — Jolly-Scientist-2484 · 2026-10-22