fix(flows): warn (not block) on inference-provider readiness while authoring, fail runs cleanly (B45) - #5212

Merged

graycyrus merged 12 commits into

tinyhumansai:mainfrom

graycyrus:fix/flows-inference-readiness-gate

Jul 27, 2026

Commits on Jul 27, 2026

feat(inference): add managed-backend readiness probe (issue B45, 1/5)
graycyrus
committed
feat(flows): add inference-readiness author gate + proposal status (B45, 2-3/5)
graycyrus
committed
docs(flows): teach workflow_builder about inference-provider readiness (B45, 4/5)
graycyrus
committed
test(flows): cover the inference-readiness gate and proposal status (B45, 5/5)
graycyrus
committed
fix(flows): stop hard-blocking authoring on inference readiness (B45)
graycyrus
committed
fix(flows): add run-time inference-readiness preflight (B45)
graycyrus
committed
fix(flows): cache a negative inference-readiness probe result too (B45)
graycyrus
committed
docs(flows): teach workflow_builder inference readiness is advisory (B45)
graycyrus
committed
test(flows): cover the non-blocking inference-readiness contract (B45)
graycyrus
committed
fix(flows): factor out inference-probe-cache type aliases (clippy::type-complexity)
graycyrus
committed
fix(flows): probe every distinct agent-node provider role (B45 findings A+B, C)
graycyrus
committed
fix(inference): tighten provider-not-configured classifier to its documented signal (B45 finding D)
graycyrus
committed