Firm Capacity / inspectable work
A report the reviewer can inspect
Decision brief
Northbank fictional diagnostic NB-EX-02 · authored output specimen · 26 September 2026 ·version 1.0. Prepared from the included I-01–I-05 teaching sources. It demonstrates substantial report preparation and a review format; it is not an executed workflow, professional endorsement, accepted client report or measured saving.
Question: where is documented handoff and completion evidence insufficient for the next owner? Recommended next decision: repair the required-field register and compare source-linked report preparation against that repaired method. No staffing recommendation, invoice approval or reuse permission follows.
Three findings: three of eight handoffs lack a recorded receipt; three record a return reason; case E lacks documented milestone acceptance. Treat receipt gaps as evidence gaps, not proof work failed. Investigate returns before assigning a cause. Finance and rights owners retain their decisions.
Findings and trace
| Finding | Supported observation | Source and limit | Proposed next owner |
|---|---|---|---|
| F-01: missing receipt | B, E and H have no receipt recorded:3/8, 37.5% of this constructed sample | I-02 rows B/E/H; missing record does not prove non-receipt; no population rate | Operations requests actual receipts or records unknown |
| F-02: returned work | A, D, G record returns for locator, scope and inference problems:3/8 | I-02 return fields; no labour, severity or causal data; not a saving estimate | Engagement lead checks source/version completeness; reviewer inspects inference |
| F-03: close-out gap | E has draft invoice reference without acceptance locator; neither C nor E has collection evidence | I-05; issue/payment status unknown; contract eligibility unverified | Client sponsor clarifies acceptance; finance decides milestone/invoice and records collection |
Implication to test: required fields and a versioned source-linked review pack may reduce reconstruction. The current records cannot establish how much effort changes. Record full preparation/review/correction/coordination on comparable cases before claiming capacity.
Alternative explanation: people may have received/accepted work without recording it. The problem may be record completeness rather than delayed work. A better incumbent register could solve most of it without AI.
Review disposition and revised handover
The following dispositions are authored examples, not an independent practitioner's approval.
| Draft assertion | Fictional disposition | Revised output and next step |
|---|---|---|
| 'Cut two leadership roles' | Reject: no workload/staffing evidence and outside I-01 scope | No staffing conclusion; partner decides any separate scope request |
| 'E invoice ready' | Hold: no acceptance locator or verified milestone | Finance receives a gap record, not an approved invoice |
| 'Reuse C report' | Reject permission assumption | Practice lead checks actual rights; default no reuse |
| 'B/E/H prove delayed delivery' | Revise inference | Three missing receipt records; operations requests evidence; delay status unknown |
Proposed handover record: case NB-EX-02; sources I-01–I-05v1.0; findings F-01–F-03; professional disposition pending in real use; customer acceptance absent; finance action pending evidence; reuse blocked. Receiving owner records its actual accept/return with exact version before state advances.
Action candidates and stop conditions
1. Operations standardises required receipt, scope version and source-locator fields in existing tools. Stop if the proposed fields duplicate an already reliable source of truth. 2. Engagement lead supplies a source-linked draft to a qualified reviewer. Compare repaired method versus AI preparation on matched cases. Stop if total effort or material errors worsen. 3. Finance and practice lead maintain separate completion and rights decisions. No invoice issue or reuse occurs merely because the report is polished.
What improved in this specimen: observation, inference and action are separated; source locators and uncertainty are visible; each gap has an owner. What remains unproven: actual effort reduction, professional quality, execution controls, acceptance, adoption and money. Original inputs and the complete Blueprint are included in the evidence pack.
Open report-inputs.html to inspect the original sources. Follow sample-blueprint.html for the complete three-design recommendation, or buyer-approval-pack.html to evaluate the purchase boundary.