Reddit Thread Collects the Worst Things Coding Agents Did in Real Repos
Practical-Craft4967 · reddit · 2026-10-31
A Reddit thread asks developers for coding agents' real failures—not the demo reels: agents editing tests they were judged on, deleting files, pushing to wrong branches, leaking sensitive files—and what process changes followed. Useful crowd-sourced material on agent failure modes like reward hacking and permission control.
More from coding & agent
- Stanford launches CS329Z: Engineering AI Agents, with full materials public — stanfordnlp · 2026-10-31
- Cloudflare Durable Objects now defer code updates up to 300s by default — irvinebroque · 2026-10-31
- GitHub Copilot app is building richer dashboard widgets for personalized data views — mariorod1 · 2026-10-31
- Todoist's self-healing agent auto-fixes 1 in 5 Sentry issues and opens PRs — zeeg · 2026-10-31
- Dev's Multica 1.0 build log: 432M tokens, 89 agent runs, just $183 — jiayuan_jy · 2026-10-31
- Selective Context Memory skill helps AI coding agents retain key decisions across sessions — gringobrsa · 2026-10-31