L15 DIAGNOSIS LOG
==========================================================================
Diagnosis: No AI.   Repair: AI allowed - keep a record (the record is at
the end; fill it in only if an assistant helped with a repair).

One log entry per failing test. Write the hypothesis BEFORE you change
any code. Make ONE change per hypothesis, then re-run ALL your tests.
Copy the ENTRY block for each further failure.

Date:                       Function:

--------------------------------------------------------------------------
ENTRY 1
--------------------------------------------------------------------------
Failing test (label, input, expected, actual):


Category from L14 (specification gap / generation error / test error):


Where (the line the traceback names, or the line you traced to):


Hypothesis (one sentence: "I think ___ because ___"):


Evidence (what you ran or printed, and what it showed; paste real output):


Hypothesis ruled IN or OUT?
  If OUT, write the next hypothesis below and gather new evidence.


Change made (exactly one; quote the line before and after):
  before:
  after:

Re-test result (paste the summary line; ALL earlier tests, not just this one):


--------------------------------------------------------------------------
ENTRY 2   (copy this block for each further failure)
--------------------------------------------------------------------------
Failing test:

Category:

Where:

Hypothesis:

Evidence:

Ruled IN / OUT:

Change made:
  before:
  after:

Re-test result:

--------------------------------------------------------------------------
AI-USE RECORD (only if an assistant helped with a repair)
--------------------------------------------------------------------------
Label:            AI allowed - keep a record
Assistant:        (name and version, or "local model")
Date(s):
Prompts:          (paste the request you sent, verbatim, or a link)
Your written diagnosis was complete before you asked:  YES / NO
What I kept:      (which of its changes you accepted)
What I changed:   (what you altered or rejected, and why)
What I verified:  (the re-test summary line, and whether the assistant
                   changed anything you did not ask it to change)
