Default Model for Scheduled Tasks

Preflight Checklist

• [x] I have searched existing requests and this feature hasn't been requested yet
• [x] This is a single feature request (not multiple features)

Problem Statement

Title: Scheduled tasks should fall back to the next-best available model when the pinned model is exhausted or retired

Problem: A scheduled task set to "Default Model" fails when that model (Fable 5) is exhausted on tokens / removed. Fails to run once that model's token allowance is exhausted — and, presumably, when the model is retired. The run produces no output, and the model can't be changed on the existing task, so the automation silently breaks until the user notices and rebuilds it.

Current behaviour: Pinned model unavailable for a run → the scheduled run stops/fails, no output, no fallback, no notification of why.

Proposed Solution

Requested behaviour: When the pinned model is unavailable for a given run (token exhaustion, model retirement, or outage), automatically fall back to the next-best available model the user's plan supports and complete the run — ideally noting in the run record that a fallback model was substituted. A per-task "allow model fallback" toggle would be the ideal control.

Why it matters: Scheduled tasks are set-and-forget automations; silent failure defeats their purpose. Model retirement (Fable 5 is a live example) guarantees every task pinned to a sunset model breaks. Graceful degradation to the next-best model preserves continuity, which is the whole point of scheduling.

Alternative Solutions

Don't have models that exhaust credits at differential rate to overall token pool

Priority

Critical - Blocking my work

Feature Category

CLI commands and flags

Use Case Example

_No response_

Additional Context

_No response_