C.26.2:4.10 - Case bank and near misses

Case	Supported C.26.2 reading	Near miss or neighboring-pattern handoff
Incident release freeze	Teams stop releases, prepare rollback evidence, redirect support work, and change escalation before a written decision appears.	If a manager issued a clear command and the work merely followed it, use work and authority patterns with evidence, not EDSE.
Service-mesh throttling regime	Route changes, saturation traces, retry patterns, and on-call routines show a coordinated throttling state under policy P.	If one controller rule fully explains the behavior, state the rule and use ordinary system/dynamics evidence.
Market expectation shift	Pricing, support inquiries, partner messages, and risk notes move together after a public signal.	If media amplification or common stimulus explains the movement, keep only that propagation/evidence claim.
Team “already decided” posture	Backlog changes, review comments, and role assignments show that the team is acting under an unstated decision.	If the claim is only a vibe or a few statements, do not use EDSE; gather carriers or keep it as informal observation.
Survey of culture	Survey answers conflict, but work traces show a stable escalation habit.	Treat the survey as a probe and state any export loss in its answers; do not make those answers the state itself.
Dashboard-following organization	Teams coordinate around the public metric after the metric becomes visible.	Apply `C.26.1` if the metric output is being used as a passive read despite probe-caused change; EDSE may carry only the residual coordinated-state reading.