Codex Desktop Bug: Missing Secret-Manager Capabilities Cause Auth Loops
jyongchul · ghdev · 2026-08-09
A bug report for OpenAI's Codex Desktop on Windows reveals missing runtime capabilities for credential consumption. Even when a task is explicitly authorized to use host-local secret managers (like Google Password Manager) for external web logins, the active session fails to expose the scoped credential tools or declared ownerassistance tools.
Because this limitation is only discovered after work begins, Codex repeatedly falls back to short owner-login windows. These windows expire, causing the external task to stall in an infinite loop. The reporter urges for a safe, auditable credential-consumption API and earlier surfacing of security boundaries or missing capabilities.
More from coding & agent
- DIRECT: retrieval-based next-question suggestions for conversational agents — burkov · 2026-10-16
- Claude Code and Codex now autocomplete the user's next action — tokenbender · 2026-10-16
- Hermes Agent ships official Reverse Engineer Anything plugin with one-command setup — Teknium · 2026-10-16
- MLExpert: 35B MoE Agent Trained to Autonomously Handle Machine Learning Engineering — Shangheng Du · 2026-10-16
- Cowork-Gym: Alibaba Synthesizes 64.9K Agent Work Environments, Boosting Qwen by 8.7 Points — Anni Zou · 2026-10-16
- MMSearch-OPD: Joint RL and On-Policy Distillation for Long-Horizon Multimodal Search Agents — PekingUniversity · 2026-10-16