Developer Reflection: Highly Autonomous Agents Are Untrusted; Approval Should Sit Before Each Action
Deep_Ad1959 · reddit · 2026-07-23
A developer on Reddit shared their reflections on autonomous agents, arguing that as agents become more autonomous, they become too untrusted to manage real accounts like Gmail or CRMs.
Instead of full autonomy, the author advocates for an agent that pauses and asks for permission right before executing a specific action. While this sounds like a downgrade, it places the approval gate exactly where mistakes would happen, making the agent safe enough to run in a live environment. The core concern isn't the agent's capability, but the risk of unsupervised actions.
Related event: The Dilemma of AI Agents: Autonomy vs. Permission Boundaries(6 posts)→
More from coding & agent
- Claude Code + Monid MCP builds a $12K/year B2B prospecting engine on live data — aryanXmahajan · 2026-10-26
- Karpathy shares CLAUDE.md rulebook to stop AI coding agents from piling up tech debt — mdancho84 · 2026-10-26
- Rethinking dependencies in the agent era: drop libs that save 20 lines of code — gethackteam · 2026-10-26
- RustGIS hits 2,000 GitHub stars in two weeks with agent-first architecture — giswqs · 2026-10-26
- Vercel Ships 12 New Products at Its 2026 Conference as It Races Rivals for the Agent Entry Point — FinanceYF5 · 2026-10-26
- Build an AI agent from scratch in under 50 lines of pure Python: ReAct loop tutorial — adnan_hashmi · 2026-10-26