Problem

Provider readiness cannot rely only on a manual refresh. The app needs background awareness of stale versions and broken auth state.

Scope

• Track non-blocking polling or refresh behavior for provider health and version state
• Define how warnings surface in the Toolchain Center and the wider workbench
• Cover operator-friendly stale-state notifications

Out of scope

• Detailed telemetry pipelines for provider health beyond product-facing state updates
• Session-execution retry behavior

Implementation notes

• Keep background checks lightweight and operator-readable
• Ensure warnings do not become hidden background noise
• Align surfaced states with the Toolchain Center model

Definition of Done

• The issue defines what should be polled, when warnings surface, and how stale state is represented
• The background readiness model is specific enough for later implementation

Verification

• Review the issue against Toolchain Center UI and trust-control backlog items

Dependencies

• Parent epic: [Epic] Build the provider toolchain center for Codex, Claude Code, and GitHub Copilot #14