M07 SELF-CHECK

Tests, review, and handoff

Five questions · 4/5 to pass · unlimited retry

1. Why compile the same policy implementation for host tests and firmware? To make host tests prove electrical correctness. To exercise the actual shared logic and avoid a second implementation drifting away. To eliminate every difference between host and board.

2. What makes an intentional mutation a useful behavior-test demonstration? The altered source compiles, the intended assertion rejects its behavior, and the good source still passes. Any compiler error occurs after an edit. The test expected value is changed to match the mutation.

3. A handoff says tests passed, but the new session cannot find a command, source identity, or result. What should happen? Repeat the passing claim because the handoff is authoritative. Implement the capstone to compensate for the missing record. Find the receipt or correct the unsupported claim and execute the relevant check when possible.

4. A consistent local-variable rename is submitted to the review skill. What is an acceptable outcome? No supported correctness finding, with any material verification limitation stated. At least one serious finding, because every skill run must find a defect. A full rewrite of the module without a requested repair.

5. The requirement keeps a sample current at age 3000 ms. A diff changes the expiration test from age > 3000 to age >= 3000. Which observation rejects that diff? A malformed fixture exits 2. The equality assertion fails because the sample becomes stale one millisecond too early. The program reports ALERT at 28 degrees in CP07.

Your result

Not attempted

Read the correct answers and explanations

Question 1

To exercise the actual shared logic and avoid a second implementation drifting away.

Shared implementation reduces drift between what is tested and what is cross-compiled. Host execution still has distinct timing, adapter, and physical limits.

Question 2

The altered source compiles, the intended assertion rejects its behavior, and the good source still passes.

A meaningful compiling defect tests the assertion rather than compiler syntax checking. Keeping the independent requirement and a passing good baseline makes the comparison interpretable.

Question 3

Find the receipt or correct the unsupported claim and execute the relevant check when possible.

A handoff points to evidence; it cannot manufacture it. Later independent execution must be labeled as a new result rather than reconstructed history.

Question 4

No supported correctness finding, with any material verification limitation stated.

Correct restraint matters. The procedure should report actionable supported defects and respect review scope, not invent issues to fill a checklist.

Question 5

The equality assertion fails because the sample becomes stale one millisecond too early.

The changed comparison wrongly expires exactly 3000 ms. The equality test follows from the requirement; CP07 has no capstone alert behavior and fixture grammar is a separate concern.