Review file delivered

Per-part completeness assessment with a status per part

Judgement on how the morning was spent

Uncovered ground identified from the feature summary

A verdict with one priority fix

A bounded next assignment for Wednesday

MUST NOT endorse bug count as the measure of the work

Constraints respected

Plan file delivered

A small named combination set, justified by the money risk

Untried combinations named, with the exposure stated

One bounded objective per block, naming what the gate needs to know

Variation beyond the flag settings

Records separate defects from behaviour needing a commercial ruling

MUST NOT leave findings unread until the gate

Out-of-scope boundary respected

Onboarding file delivered

The afternoon is split into bounded stretches, none of them a marathon

One objective per stretch, stating what she should be able to answer

The first stretch is orientation and it aims the later ones

Notes separate product knowledge, suspected defects, and questions

A 17:00 hand-off with a stated fate

MUST NOT propose one unbroken multi-hour sitting

Off-limits areas named

Plan file delivered

The two clock jumps are budgeted as the scarce resource

The propagation dead time is used

A single bounded objective naming what must be learned

Untouched areas named

Records separate defects, unexplained device behaviour, and rig problems

MUST NOT leave the record in a chat thread

Rig constraints respected

Plan file delivered

A single bounded assignment naming what must be learned

Fits one 90-minute window with named untouched areas

Patient-safety risk drives the assignment

MUST NOT prescribe an ordered script - seeds, not steps

Recording structure separates behaviour, defects, and blockers

Closing write-up with coverage gaps and a tester judgement

Named reader and a deadline for the write-up

MUST NOT expand beyond W-TEST or into the mobile app

Report file delivered

MUST NOT attempt all eleven angles - the list is cut and the cut is justified

A single stated objective for the hour

Sized to sixty minutes with time actually allocated

Concrete attack ideas on the kept angles

Findings split between defects and rulings needed

Gate report carries coverage gaps and a tester judgement

Named next target if the gate is held

MUST NOT plan rooted or jailbroken device work

Campaign file delivered

A ranked backlog of bounded pieces, longer than the budget

Six blocks allocated, with the remainder explicitly dropped

The plan adapts mid-campaign

Per-block record and hand-back

MUST NOT produce an aggregate nobody acts on

Stopping and escalation condition

Payroll-core ownership boundary respected

Plan file delivered

Prior defect history used as the standard of comparison

One bounded objective per block naming what must be learned

Coverage beyond the historical list

Records separate regressions, acceptable differences, and questions

MUST NOT hand back untraceable findings or an open-ended continuation

Time is allocated within each block

Flagged work stays out

Plan file delivered

Per-block time accounting with named categories

A trustworthiness threshold and an agreed rule for a bad block

One bounded objective per block, naming what is to be learned

Records separate app defects, lab problems, and decisions needed

Per-block hand-back including coverage gaps and a qualitative read

MUST NOT leave the time breakdown unused

3PL boundary respected

Plan file delivered

One bounded objective per block, stating what is to be learned

The 90 minutes are actually allocated and the plan is sized to fit

Store-credit mismatch is prioritised over breadth

In-flight notes separate confirmed behaviour, defects, and open questions

Per-block hand-back that closes the loop

Stopping condition and time accounting

MUST NOT bundle both objectives into one block or exceed the block length

Out-of-scope boundary respected