Where Coding Agents Should Ask for Confirmation
Substantial-Heat-321 · reddit · 2026-07-16
The author discusses what control interfaces like Codex Micro should confirm before an agent executes code. The core isn't a simple "yes/no," but making the **scope, cost, and irreversible actions** visible at the right moments. Three checkpoints are proposed: - **Before starting**: Confirm the repository/branch, accessible paths, runnable commands, and token/time budget. - **During execution**: Display concise progress, including plans, modified files, test status, and whether the task has exceeded its original scope. - **Before irreversible actions**: Require explicit confirmation for unauthorized writes, destructive commands, credential access, external calls, and PRs/merges, while showing the specific action, affected resources, and current evidence. Finally, the author asks: Which of these controls have actually helped you avoid costly mistakes? And what additional information do you wish the interface displayed before approving the next step?
More from coding & agent
- Autoresearch proposes packaging ML runs as studies with questions, analysis, and code diffs — morgymcg · 2026-07-21
- CHAP defines approvals, handoffs, and audit logs for human-agent workflows — DeliveryTechnical199 · 2026-07-21
- The author says Codex reached 20x and is now debugging spec decoding on a hybrid parallel setup — TheZachMueller · 2026-07-21
- Axcess adds an MCP connector for WCAG accessibility checks that scanners miss — modelcontextprotocol · 2026-07-21
- X post asks whether Cursor Composer, built on Kimi models, would also be banned — max_paperclips · 2026-07-21
- A developer’s Codex usage is draining pooled enterprise credits at a small company — Distinct_Relation_62 · 2026-07-21