Repository navigation
Configurable fallback chain when a running thread hits a usage limit #34285
mussaddiqmahmood7
started this conversation in
Ideas
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Problem
fallbackModelSelectionis a single slot and only applies when a new thread starts. When a running thread hits a usage limit, recovery waits for the reset time or needs a manual model switch. With several Codex accounts plus another provider, an unattended thread just stalls.Proposal: user-controlled, opt-in
Nothing automatic for everyone. Users define their own ordered fallback list in Settings > Providers/Models. Each entry is a provider instance + model + effort/options, chosen explicitly, since effort levels are not equivalent across providers (e.g. Codex "High" vs Claude "Medium" is the user's mapping, never inferred).
Example: Codex account 1 (High) -> Codex account 2 (High) -> Claude Opus (Medium).
When a run stops on a usage limit and the feature is enabled:
Scope
Related
#28226, #18721, #32188
Happy to help implement if maintainers approve the direction.
All reactions