Rubric and Readiness

Score each dimension 0 (missing), 1 (works with substantial help), or 2 (independent evidence).

• Behavior: parsing, summaries and persistence meet the specified boundaries and error cases.
• Testing: expected results are independently predicted, and a regression test catches an observed or explicitly seeded defect. The diagnosis note distinguishes hypotheses using an experiment or trace.
• Explanation: accurately trace values, calls, data choices and the storage boundary.
• Transfer: independently satisfy the original reading-page contract and the new first-over-budget clinic task, then explain the changed requirement. Close the delayed checks in the pilot record.
• Evidence: another learner can run the project and inspect the debugging note and assistance declaration.

Pass only with 2 in every dimension (10/10). Strength in one category cannot compensate for missing testing, explanation or evidence. A destructive corrupt-file fallback is an automatic repair requirement. Passing a checker without explaining the code is not a pass. Self-assessment is provisional; independent review and delayed retention evidence increase confidence.

If behavior is weak, repair the smallest failing contract. If explanation is weak, return to tracing. If transfer is weak, build a smaller fresh example rather than rereading the solution. Reattempt the failed dimension with a new input and preserve both results.

Immediate evidence can meet each dimension’s immediate requirements while delayed Transfer evidence remains pending; do not record a final 10/10 until both delayed checks and their repairs close. Immediate success permits provisional progression under the pilot and retention rules. Record independent review as pending if unavailable. On passing, proceed to Module 1 and add the lab evidence to the Semester 00 checkpoint.