Simulation: FrontierOrg is a fictional infrastructure holding group; figures are fixtures or marked not measured, and no vendor sponsors this work.

Baseline 3 October 2026 · latest scenario note 19 November 2026 (Day 34) · files in scorecards/

Empty on purpose.

Three scorecards, linked to the same scenario. A fixture row count is not a result. “Not measured” is the honest actual. Day 34 continues revise, on hold: stock fact has no source (DATA-004); a desk sample flagged 2 candidate safety false negatives on the parts rows (RISK-002), and the human-only list stays unchanged until a named safety reviewer samples it. Day 26 maps a CRM note as a paper candidate that cannot change the promise date. Day 27 writes the held-invoice / billing assist as a blocked card (hold release stays human-only; billing assist mapped as pilot false). Day 28 writes the duplicate business-partner merge as a blocked card (no silent merge; fixture unrepaired). Day 29 red-teams the CRM note on paper and keeps it a paper candidate. Day 30 folds the four bounds into the draft-only design on paper (v2, proposed, not adopted; reviewer seat unfilled). Day 31 freezes new workload mapping and inventories where the draft would wait (7 stages, no named store or retention). Days 32–34 hold with no new artefacts. On publish (26 Oct 2026): Craig declined to name an outside reviewer (26 Oct 2026). Reviewer seat: Declined by sponsor, 26 Oct 2026; RISK-003 open; the mapping freeze lifts and DATA-004 ownership on paper is next. Every boundary is unreviewed author judgement. Decision text, not an actual.

C-suite
IDQuestionBaselineActualDecision
CS-CUSTOMERCustomer outcome on the traced promisenot measurednot measuredSend/act no-go. Draft-only revise-candidate. Stock-check build paused. Specification and goods-in checks on parts stay with people; 2 parts tasks mapped as blocked workloads, not pilots. CRM note on paper only; cannot change the promise, stock flag, or invoice hold. Billing hold assist blocked as pilot; hold release stays human-only. Duplicate BP merge blocked; no silent merge. CRM note must not state or imply the restoration date (v2 bound proposed, paper). Day 30: bounds folded into AGT-61003-DRAFT v2 on paper; only output renamed internal status note; stock-check gate kept. Day 31: proposed bound B keeps the contact's name and email out of the prompt (paper).
CS-PEOPLEEmployee transitionnot measurednot measuredStores and parts planner and field supervisor: roles only, unnamed, unconsulted. No fictional safety lead invented. No fictional finance control owner or master-data owner invented. Amina in view for the CRM red-team; v2 bounds unseen by anyone doing the work. Day 30: Amina keeps the date, the record, and the send; reviewer outside the fiction unfilled, none invented. Day 31: Amina owns 2 of 7 draft lifecycle stages; no privacy lead invented. Days 32–34: hold; nobody asked to read anything new. Reviewer seat: Declined by sponsor, 26 Oct 2026; none invented. Workforce sessions: 0. No task removed.
CS-RISK-COSTRisk and cost controlnot measurednot measuredRecommend revise. Envelope empty. Assistant not deployed. RISK-002: 2 candidate false negatives, human-only list not grown, no named reviewer. Billing hold and duplicate BP merge boundaries documented as blocked. CRM draft persistence and retention unnamed (RISK-001); agent executions 0. Day 30: v2 lists 11 forbidden actions (was 7), loosens 0 controls; transient store and logging unnamed. Day 31: mapping frozen; 7 stages, named store 0, retention 0; 3 unpriced cost lines. Days 32–34: hold; 0 new artefacts. 26 Oct 2026: reviewer seat declined by sponsor, 26 oct 2026; RISK-003 open (no review took place); mapping freeze lifts.
Operational
IDNameBaselineActual
OP-CASESCases in the published fixture1 fixture rownot a rate
OP-CYCLECycle timenot measurednot measured
OP-EXCEPTIONExceptionPromise before stock; invoice heldnot measured
OP-ESCALATIONEscalationsnot measurednot measured
OP-QAQuality assurancenot measurednot measured
OP-TOKENSToken or credit usenot measurednot measured
OP-ROLLBACKRollbackNothing deployednot measured
Operations area — Northwater field service
IDNameBaselineActual
BA-HOURSObserved hoursnot measurednot measured
BA-TRAININGTrainingNot startednot measured
BA-QUALITYQualitynot measurednot measured
BA-COSTCostsnot measurednot measured
BA-BENEFITBenefitsNot claimednot measured
BA-CONTROLControlHumans named; no agent in the pathnot measured

Days 13–34 add an outcomes register, an empty cost envelope, a licence chase without invented regions, a stock-check design (fixture still defective), a draft-only agent design, a tabletop red team, handoffs, portfolio ranking, a go/revise/stop pack, Day 23's finding that the stock flag has no source (DATA-004), Day 24's parts safety desk sample (labelled counts: 8 rows sampled, 2 candidate false negatives, 0 labels changed, named safety reviewer none), Day 25's blocked parts workload cards (labelled counts: 2 blocked candidate workloads documented, human-only pattern list grown false, 0 safety labels changed, named safety reviewer none), Day 26's CRM note paper candidate (labelled counts: CRM note candidate documented true, billing assist mapped as pilot false), Day 27's blocked billing hold card (labelled counts: 3 blocked candidate workloads documented, 4 blocked-or-paper candidate workloads documented, billing hold boundary documented true, billing assist mapped as pilot false), and Day 28's blocked merge card (labelled counts: 4 blocked candidate workloads documented, 5 blocked-or-paper candidate workloads documented, duplicate BP merge boundary documented true), and Day 29's paper red-team of the CRM note (labelled counts: 6 stories, 3 wording gaps, 1 partial, 4 v2 bounds proposed, agent executions 0; findings about wording, not a pass rate), and Day 30's v2 draft-only design on paper (labelled counts: 4 bounds folded, 11 forbidden actions, 1 O*NET task in the draft's scope, 0 controls loosened, 0 new workload cards, named external reviewer none, agent executions 0), and Day 31's draft lifecycle inventory under a mapping freeze (labelled counts: 7 stages, 0 named stores, 0 retention periods named, 2 v2.1 bounds proposed, 0 folded, 3 unpriced cost lines, 0 new workload cards, agent executions 0), and Days 32–34's hold (labelled counts: 0 new artefacts, hold days 3), and the 26 Oct 2026 sponsor decision (labelled counts: named external reviewer none — declined by sponsor, external reviews completed 0, workload mapping frozen false, RISK-003 open). Fixture r2 (data/sample_r2/, 1 parts reservation, 1 unnamed role) sits alongside r1; it is a fixture, not a result. None of that moves an actual. Training asks are not a course. Cost envelope amounts measured: 0. Forecast range: not modelled. Cost to date: not measured. Sponsor names are fictional.

History: scorecards/history/day-01.json through day-34.json. Standing issues remain open.