Dev runs Codex as a shell tool inside Claude for code review, skipping tmux hassles
abhi9u · x · 2026-10-14
Developer abhi9u shares his dual-model coding workflow:
- Split: Claude writes the code, Codex advises and reviews — he finds Codex "super in finding holes."
- Old pain: He first had Claude talk to Codex via tmux commands, but approvals stalled — Codex would wait for permission while Claude never checked, wasting lots of time.
- New setup: Claude now executes Codex as one of its shells and treats it like a tool, which is far more effective.
A practical multi-agent pattern: have your primary coding agent orchestrate a reviewer model as a subprocess.
More from coding & agent
- Startup launches cloud computers built for AI agents: 4 vCPUs in under a second — CurieuxExplorer · 2026-10-15
- User has Grok bot build its own email UI, and it just works — elonmusk · 2026-10-15
- YC F26's Papaya launches to auto-fix production AI agents via PRs — ycombinator · 2026-10-15
- Eval-driven development for RAG: keep eval cases in Git, never regress — Al_Grigor · 2026-10-15
- LlamaIndex launches OpenDocRouter with ParseBench benchmark comparisons — llama_index · 2026-10-15
- User reverse-engineers Grok's X news curation prompt and turns it into a reusable agent skill — dotey · 2026-10-15