Dev shares a tiered file-permission policy to stop agents stalling on approvals
gaganghotra_ · x · 2026-11-18
Responding to the common problem of AI agents stalling at "allow permissions" prompts, a developer shared a practical authorization policy for file operations:
- Create file: just do it, no asking.
- Delete file: allow it, but keep a copy in a designated folder and log why it was deleted.
- Update file: allow it, but keep a copy of the old version in a designated folder and log what changed and why.
The core idea: allow the action but always keep backups and an audit trail, replacing per-step human confirmation to stop agents from idling.
More from coding & agent
- YC-backed Ample launches one-stop cloud for agents, hits 5,000 users two weeks after launch — ycombinator · 2026-11-18
- OmniLoRA: train any model on any GPU inside your ComfyUI graph — pente5 · 2026-11-18
- findatasets ships /feedback endpoint so AI agents can file bug reports, following Coinbase's lead — virattt · 2026-11-18
- Put rules in code, not prose: what actually makes an AI instruction a Rule — gerardsans · 2026-11-18
- Museo, a whisperable AI ring for phone-free walks, ships next month with open dev platform — Scobleizer · 2026-11-18
- Corenna MCP server builds signed Android/iOS apps from GitHub repos inside ChatGPT or Claude — Personal_You3422 · 2026-11-18