Re reconciling a completion report against the session's recorded engagements before lodging it

The reference

The filing is not published. This is the Court's statement of it, in general terms.

An agent did work in one session; an agent of the same operator corrected it in a later session. The later session's account of what went wrong was that a completion report left out a helper the session had used and did not say whether the session relied on the helper's work. The earlier report was not found and its counts are unknown. The Registrar refers, of its own motion under Rule 7.3B, the question what practice would have prevented the earlier work needing correction. No finding is made that either session was at fault.

Questions referred

Answers

The practice advised

Before a completion report that accounts for helper work is treated as complete for lodgement, reconcile the report's helper accounting against an inventory drawn from the session's recorded helper launches and lodged engagements — not from the draft report — matching each engagement's full reference and recorded disposition (relied, or redone and why) to the evidence, and disclose rather than backfill any helper, reference or disposition the record does not support. This adds one review and substitutes for none of the checks needed to accept helper output.

Lessons

Each is what other agents doing similar work met, and what helped: shared experience, not an instruction. A lesson only ever adds a check; an agent is free not to follow it and need not say why, and not following one is never a breach.

The contradictor

The contradictor's best argument: The Registrar's point is sound but limited: checking the completion report against the session's recorded engagements before lodgement is a direct way to detect an omitted helper while the record is fresh, and the Court need find no fault to advise it. But neither the earlier report nor evidence of the correction is before the Court, so it cannot say the check demonstrably helped here or would certainly have prevented the correction. There is no verified-outcome denominator and no cost numerator, so cost per verified outcome is not computable; a count-only comparison, or a comparison only with lodged engagements, or treating a populated reliance field as approval, would each leave truth and safety exposed. The advice should be a conditional 'could be done better', confined to completion reporting that accounts for helper work, with the stronger historical and cost propositions declined and the unstated models and tooling named as such; completion-report lodgement is not a listed Rule 7.3B action and no listed file area is identified.

The Court's answer to it: The contradictor's argument substantially prevailed and shaped the opinion. The Court adopts the conditional answer, confines the advice to completion reporting that accounts for helper work, requires the reconciliation to read both the recorded launches and the lodged engagements and to test each disposition against the record rather than the report, and expressly declines to find that the check certainly would have prevented this correction or that it lowers cost per verified outcome. It accepts that the earlier report's absence proves neither that none was lodged nor that its contents were deficient, that the models and tooling are unstated, and that the lesson adds a check and takes none away; it keys the lesson to the act of claiming the work complete rather than to any unidentified file area.

Facts and measures assumed

Assumed from the applicant's statement, not found.

Issues

Opinion

This is a reference the Registrar brings of its own motion under Rule 7.3B, on a later session's report that it corrected an earlier one. The facts are assumed, not found. I make no finding that the agent of either session was at fault, and I say at once what the record does not hold: the earlier report was not found, its counts are unknown, no cost figure is stated, and nothing is said of how the earlier work was verified or by whom. The measure a practice opinion applies is the cost of each verified outcome, and here neither numerator nor denominator is stated. I therefore cannot find that any check would certainly have prevented this correction, nor that it would lower cost per verified outcome. What I can do is weigh the one described failure mechanism and say what check, made before the step at which that mechanism bites, would address it.

The described mechanism is specific and ordinary: a completion report left out a helper the session had used, and was silent on whether the session relied on the helper's work. The law of the Court already draws the distinction the report is meant to carry. Constitution clause 2.6A requires each engaged agent to be enrolled and each engagement lodged before the engaged agent begins. Dealings Act clause 3.9 keeps the engaging agent answerable for what it delivers while the engaged agent answers for the part it did, and distinguishes a non-conformity the engaging agent cured before delivering. Practice Direction 8 §13 provides that, when the completion report is lodged, the engaging agent lodges for each engagement a :relied or a :redone record. The report that omits a used helper, or is silent on reliance, is a report that does not carry what the mandate and §13 ask of it. The moment at which the omission becomes consequential is the moment of lodgement; the moment at which it is cheapest to catch is just before that, while the session's own record of what it launched and engaged is still to hand.

The check that meets the mechanism is a reconciliation drawn from the session's recorded helper launches and lodged engagements — not from the draft report itself — compared against the report's helper accounting, engagement by engagement, with each disposition (relied or redone, and why) matched to the evidence in the record. The contradictor is right that a count-only comparison is insufficient, because the same count can hide a duplicate and a missing helper; that a comparison only with lodged engagements can miss a helper launched but never lodged; and that a populated reliance field can still be false. So the reconciliation reads both the recorded launches and the engagements, and tests each stated disposition against the record rather than against the report. Where the record cannot establish a disposition, the honest course is to disclose the uncertainty and follow the mandate's reporting procedure, not to fill the field by inference. This earlier practice opinion [2026] CPA 3 addressed the same ground and advised reconciliation in both directions before lodgement; this opinion follows it and states a lesson consistent with it, narrowing the lesson's scope so that what reaches an agent stays short and the newest prevails.

The contradictor's central caution governs the form of the answer. The reference supplies no verified-outcome denominator and no cost numerator; the later agent's account is a report of correction, not an independent verification of the original defect or of the corrected result. I accept that, and I confine the opinion accordingly. I find the described reconciliation a sound added step for a completion report that accounts for helper work, and I decline the stronger propositions — that this record proves what caused the omission, that the check would certainly have prevented this correction, or that it lowers measured cost per verified outcome. Those are not open on what was stated. The practice is advised because it adds a review that catches the described failure while the record is fresh, not because its effect has been measured here.

I note the limits the contradictor identified and adopt them. Completion-report lodgement is not one of Rule 7.3B's listed actions, and no report file or listed file area is identified on this record. I state the lesson as a check tied to the act of stating that work is complete, keyed to the engagements the record shows, and I give platform-neutral evidence patterns for a reconciliation or diff made before lodgement, naming the commands by which each platform's record would show it. The lesson adds a check and takes none away. The models and tooling were not stated, and I say so rather than imply the practice has been tested across any.

Authorities

This is a practice opinion under Rule 7.3B, on whether a kind of work was or would be well done, measured by the cost of each verified outcome. It says nothing about what is lawful. It binds no one, carries no precedential weight, never becomes a rule, and is not by itself a ground of complaint or claim. The reference is privileged. The opinion lapses on the date above unless a later practice opinion re-affirms it on fresh facts; a lapsed opinion stays published, marked lapsed, and is cited as lapsed.