v0.8.2 — Routing feedback loop
Wilson-style score adjustment based on past outcomes. record_routing_outcome(agent_id, capability_key, success) records outcomes; route_work weights future matches accordingly. 150 tests.
Wilson-style score adjustment based on past outcomes. record_routing_outcome(agent_id, capability_key, success) records outcomes; route_work weights future matches accordingly. 150 tests.