Codex Desktop Bug Incorrectly Rehydrates Terminated Subagents as Working
palmer-Eormen · ghdev · 2026-08-08
A recent bug in the OpenAI Codex desktop app incorrectly rehydrates completed and aborted subagents as Working upon application restart.
Impact & Details:
- Causes false active-agent counts, making it impossible for users to determine if background work actually exists.
- Inspected child sessions are actually terminated (backend returns notfound, SQLite state is closed), yet the UI shows them running for days.
- In one affected conversation, all 76 completed subagents were displayed as actively working.
- Reconstructing numerous historical children also introduces UI/session-loading overhead.
More from coding & agent
- Constraints you force on models today will become tech debt in months, warns dev — zack_overflow · 2026-10-16
- Liquid AI tutorial: fine-tune your d1-3B decision model in 30 minutes on a free GPU — helloiamleonie · 2026-10-16
- Alibaba open-sources OpenGCUA, a full-stack generalist computer-use agent; +19.6 points on OSWorld-v2 across 8 models — Tongyi-MAI · 2026-10-16
- BotBeat: a social network where all users are AI agents, exposed via 32 MCP tools — Dismal_Piccolo4973 · 2026-10-16
- DKGI paper: agents write their own tools to search knowledge graphs, beating baselines by 6.7 Hit@1 — yuz9yuz · 2026-10-16
- Dev open-sources lgtmaybe: running PR code reviews on a local 27B model via hunk splitting — mattjcoles · 2026-10-16