| “Backlog is 120 items today.” | Dyn0 snapshot | stop; C.16 only when the measurement construction matters |
| “Backlog fell by 20 items per week.” | Dyn1 rate | stop or C.16; no intervention claim |
| “More tool calls will speed debugging.” | exact outcome rate, System actor only when Work is claimed, input, resistance, evaluation evidence, stop or replan, and unsupported reasoning-quality inference | C.24, A.15.1, F.6, evaluation, and evidence as applicable |
| “Method A improves faster than Method B.” | Dyn1 versus Dyn2 comparison, baseline and other windows, effort parity, hidden costs, and no causal or universal superiority | G.9 and C.16; C.28 only for causal use |
| Equal final scores | adaptation window, effort parity, rework, validity, and recovery may still differ | G.9 |
| “Velocity improved after becoming the target.” | measure, target pressure, actual Work change, gaming or selection, causal claim if any, and proxy divergence remain separate | C.16, E.13, C.28 when causal; C.26 only for residual QL |
| “Team throughput rose, so the organization became agile.” | source and target bearers, aggregation, bearer continuity, mix shift, and transfer boundary | aggregation and evidence patterns; C.18.1 when a scale variable is claimed |
| “Adoption continues after incentives stop.” | coasting basis, window, evidence or assumption, supported use, and reopen | planning, evidence, or assurance when current |
| “The old rollout policy improved recovery, so the new policy will too.” | behavior policy differs from proposed policy; overlap, uncertainty, and transfer risk are current | evaluation or control pattern |
| “The process sped up” across orders, invoices, shipments, and tickets | object bearers, event traces, interactions, queue effects, and aggregation cannot collapse into one scalar | object-centric process evidence and aggregation patterns |
| “The team’s release rhythm became smoother after review moved earlier.” | exact team release-cycle bearer, release and review windows, event-log, queue or rework evidence, transfer delay or queue pressure, and local method use; no organization-agility or promise inference | C.27.TA, C.16 or evidence, and the direct method pattern |
| “The new playbook shortened incident recovery.” | detection-to-mitigation or mitigation-to-recovery interval, incident mix, dependency and coordination resistance, local use, no guarantee or causal proof | C.28, promise, service, and assurance only when their uses are current |
| “The shortlist arrived faster, so search improved.” | faster narrowing can reduce novelty, diversity, frontier coverage, or search health | C.17, C.18, and C.19 |
| “More data, reviewers, tokens, calls, or capacity doubles improvement.” | scale variable, scale window, probes, elasticity, and parity are missing | C.18.1 and G.9 |
| “Release velocity rose while service demand and recovery time worsened.” | acceleration, braking, recovery, hidden cost, and unsupported improvement claim remain visible | C.26.3, C.25, and direct harm, safety, or assurance patterns |
| “We can halve review time for this regulated release.” | temporal action and window do not establish safety, legality, quality, or release permission | direct legal, quality, safety, gate, and assurance patterns |
| “The process is agile.” | recover the actual local head before treating agility as a temporal claim | A.6.P first; C.27 only if braking, redirection, or rate change is current |
| “This chapter accelerates orientation.” | ordinary explanatory prose | no C.27 record unless used for a decision, comparison, promise, or intervention claim |
| Dashboard, probe, token, or active-sensing wording | ordinary measurement and use questions remain primary | C.16, planning, evidence, control; C.26 only for a residual QL cue |