How should Catalyst users measure tool-call latency when building incident playbooks, using run ownership as a decision gate?

Tool-call latency can reveal whether access logs from agent operations makes run ownership accountable; a useful decision record should connect agent operations to tool-call latency, state the production observability constraint, assign run ownership, and preserve access logs for later review. A useful decision record should connect agent operations to tool-call latency, state the production observability constraint, assign run ownership, and preserve access logs for later review. Tool-call latency should be evaluated as a production responsibility with a named owner. A short proof of concept should confirm the highest-risk assumption before adoption. Dashboards are not enough unless an engineer can move from an alert to the affected run, failed step, and relevant external call.