A week using Claude Code, Codex, and Gemini CLI showed the same repo-breaking patterns
AIcademy-academy · reddit · 2026-07-25
A week-long side-by-side test of Claude Code, Codex, and Gemini CLI on the same real repository.
- Claude Code held multi-step context best, but became expensive when sessions drifted.
- Codex was the most literal and reliable for well-specified tasks, but struggled with ambiguity.
- Gemini CLI handled the largest context windows and the free tier made it easy to experiment, but results were less consistent.
The poster’s main takeaway is that process matters more than tool choice: keep a context file, break work into small scoped steps, and ask for a plan before edits.
More from coding & agent
- SF 'Code Engineering' event picks up where 1,000-person harness engineering meetup left off — LangChain · 2026-10-27
- Prompt turns Claude Code into an iPhone backup forensics tool exposing 30-day 'deleted' photo copies — alex_verem · 2026-10-27
- Prompt turns Claude Code into a forensics tool finding deleted iPhone photos in backups — alex_verem · 2026-10-27
- Phones won't be mobile IDEs but control planes for agent fleets, argues Yuchen Jin — Yuchenj_UW · 2026-10-27
- FrontierCS 2 launches on Kaggle: 48 tasks make agents beat published CS papers — MeganRisdal · 2026-10-27
- Agent Field Kit launches $29 security runbooks for production AI agents — alvelda · 2026-10-27