[Dots] Cloud computer unusable: exec-server disconnected, capability query failed, no usable execution environment #52531

What issue are you seeing?

My dot's own cloud computer became unusable on October 9, 2026. Both cloud terminal commands and cloud browser operations stopped responding, then executor creation/readiness checks failed. Chat and connected-app reads, including Library, still work.

The failure was still unresolved as of 10:53 UTC. The diagnostics below were returned by the assistant's tools. I do not know the root cause.

What steps can reproduce the bug?

This is an observed incident, not a deterministic reproduction. All times below are approximate and in UTC on October 9, 2026:

1. At 09:46, cloud browser operations and terminal commands stopped returning. A cloud browser getState request subsequently failed with:
   timed out awaiting tools/call after 300s
2. At 10:02, terminal process creation failed with:
   Failed to create unified exec process: exec-server transport disconnected
3. At 10:29, the cloud environment reported failed. The diagnostic included:
   failed to query executor configuration capabilities
   The environment identifier is omitted.
4. At 10:36 and again at 10:39, minimal printf / pwd probes failed immediately, before execution:
   No usable execution environment is available. Wait for an environment to become available before using this tool.

The desktop client update had completed by 10:39, but the cloud retest still failed. The exact client version was not captured, and this does not establish any relationship between the update and the failure.

What is the expected behavior?

The dot should be able to execute cloud terminal commands and use its cloud browser, or provide an actionable failure status and a supported recovery path that preserves the existing workspace.

Please investigate the cloud environment's configuration/startup and executor connection failures, and advise how to restore access without resetting existing files and state.

Additional information

• My local Mac remained usable and successfully completed local Git operations during this incident.
• No cloud-computer restart was performed as part of these diagnostics.
• The OpenAI status page showed normal operation when checked at 10:38 UTC, with no corresponding October 9 incident listed. Aggregate status does not rule out an individual environment failure.
• No data loss is claimed; the unavailable cloud computer prevents checking its current contents.
• Client, macOS and managed-executor versions were not captured. Account details, private project information and runtime identifiers are omitted.

Potentially related reports: #50545 (capability-query and unavailable-environment errors), #50115 (cloud computer unavailable while chat remains usable), #50130 (closed as a duplicate, not evidence of a fix), and #52509 (another October 9 executor failure). A shared cause has not been verified. This report supplies the timestamps and error progression for a separate affected account.