# Restore the declared starting state between rehearsals

Illustrative planning brief; no automatic product import.

Synthetic planning case: Two checks require approval absent at their declared origin: A starts pending and B starts ready. After check A sets approval present, check B is run without restoring that flag and passes a transition that should be rejected in B’s intended fixture.

## Decision

Check B did not run its intended approval-absent fixture. Preserve its actual starting state and rerun after a verified reset before treating it as false-guard evidence.

## Owned work

- Specify the full fixture vector
  - Owner role: Rehearsal owner
  - Acceptance evidence: Starting state and approval flag are both recorded.
- Inspect setup between runs
  - Owner role: Reviewer
  - Acceptance evidence: A’s retained approval flag explains why B does not test the intended condition.
- Verify the fresh fixture
  - Owner role: Acceptance lead
  - Acceptance evidence: A new B run records the reset flag and resulting rejection/unchanged state.

## Workflow

1. Specify the full fixture vector. Check: Starting state and approval flag are both recorded.
2. Inspect setup between runs. Check: A’s retained approval flag explains why B does not test the intended condition.
3. Verify the fresh fixture. Check: A new B run records the reset flag and resulting rejection/unchanged state.

## Judgment

The invented flag model is external; a reset procedure itself needs evidence and may not be available in a real product.


## Filled manual planning note

Check B did not run its intended approval-absent fixture. Preserve its actual starting state and rerun after a verified reset before treating it as false-guard evidence. B’s requested origin state is ready in both descriptions, but its approval flag differs. Resetting a visible state label alone would not restore the full fixture vector. The invented flag model is external; a reset procedure itself needs evidence and may not be available in a real product.


## Workflow questions

### Is B necessarily a software failure?

No. The observation may match its actual true-guard input; the intended test setup was wrong.

### Can I edit the log to say approval absent?

No. Preserve actual history and add a fresh correctly established run.

## Product connection

Use the owned checks and downloaded brief to discuss this planning decision alongside your TeamBoostAI tasks. Confirm available fields, roles and account features separately. The example is manual; it does not calculate live analytics, create work or run an experiment in the product.

Confirm account availability before adopting this manual outline.

## Original worked case

Synthetic records, manual planning only. No account import or live analytics.

### Inspect the invented case records

Check | Intended start | Actual start | Observed result
--- | --- | --- | ---
A | Pending; approval absent | Pending; approval absent | Approval set present
B | Ready; approval absent | Ready; approval present from A | Accepted
B fresh | Ready; approval absent | Must verify after reset | Expected rejection

### Reasoning

B’s requested origin state is ready in both descriptions, but its approval flag differs. Resetting a visible state label alone would not restore the full fixture vector.

### Bounded result

Check B did not run its intended approval-absent fixture. Preserve its actual starting state and rerun after a verified reset before treating it as false-guard evidence.

### Distinct decision

This checks full starting-fixture restoration between rehearsals rather than general retry behavior.

### Limits

The invented flag model is external; a reset procedure itself needs evidence and may not be available in a real product.

### Definitions and method context

- NIST: state-based and event-sequence testing context — https://csrc.nist.gov/Projects/automated-combinatorial-testing-for-software — General context for state-based and event-sequence testing. All graphs, guards, fixtures and recovery contracts here are original toy definitions; the source does not certify their completeness or product behavior.