Pro account: "Selected model is at capacity" on every model, every surface, ~10 days

Thread ID: 01a0af8a-5c42-7d62-a6ef-f0bf22fab9be

Every model on my Pro account returns "Selected model is at capacity. Please try a different model." across all surfaces - Codex CLI, desktop app, Amp, Conductor, and the ChatGPT app itself. This has run for roughly 10 days, with brief recovery windows (today it worked 5:55-6:18 PM IST, then failed again).

• codex doctor --json: fully clean (auth, endpoints, websocket handshake all pass)
• Weekly quota mostly unconsumed - requests fail before consuming anything
• Remediation completed: password reset, 2FA enabled, sign-out of all devices, fresh login - zero change
• A separate third-party app on the same account (Amp) fails intermittently with "subscription Is Overloaded - this connection's provider reports that its servers are overloaded"
• Support case 15250030: specialist said access "may be temporarily limited based on account activity" with no cause, no timeline

Start with the codex doctor --json entry point and compare its clean authentication, endpoint, and websocket results with the capacity failures reported across Codex CLI and the other listed surfaces. Done means identifying why this account is rejected across models or documenting that the cause is outside the client, with a reproducible diagnosis.