L13 GENERATION REQUEST
==========================================================================
Parts 1 to 4: No AI (your prediction and your assessment are yours).
Sending the requests: AI allowed - keep a record (the record is at the end).

Fill in every part BEFORE you send anything to an assistant.
If you revise the request later, keep the old version and note what you
changed and why. Do not "chat until it works".

Task number:
Date:

--------------------------------------------------------------------------
PART 1. THE VAGUE REQUEST (copy it exactly from the task card)
--------------------------------------------------------------------------


--------------------------------------------------------------------------
PART 2. MY PREDICTION (write this before sending the vague request)
--------------------------------------------------------------------------
How will the vague request fail? Name at least two specific things the
model will have to guess, and say what you think it will guess.

1.

2.

--------------------------------------------------------------------------
PART 3. THE FULL SPECIFICATION
--------------------------------------------------------------------------
For the lab, copy the PRECISE request from the task card here. For the
homework, write your own specification in the same shape.

Function name and one-sentence purpose:


Inputs (for each: name, type, allowed values or format, one example):


Output (type, format, one example):


Constraints (Python 3 standard library only? no printing? no input()?):


Acceptance criteria (numbered; each one is a specific input and the
exact expected result, so anyone can check it; add rows if you have more):
  AC1
  AC2
  AC3
  AC4
  AC5
  AC6

Boundary behaviour (what happens exactly AT each edge):


Invalid input (what to raise or return for wrong type, empty, out of range):


--------------------------------------------------------------------------
PART 4. COMPARISON (after both outputs come back)
--------------------------------------------------------------------------
For each acceptance criterion, mark both outputs MEETS / FAILS / CAN'T TELL.
Add rows if you have more criteria.

        Vague output        Precise output
  AC1
  AC2
  AC3
  AC4
  AC5
  AC6

Was my prediction right? What did I miss?


What still went wrong with the precise output, if anything?


--------------------------------------------------------------------------
AI-USE RECORD
--------------------------------------------------------------------------
Label:            AI allowed - keep a record
Assistant:        (name and version, or "local model")
Date(s):
Prompts:          (paste every request you sent, verbatim, in order, or a
                   link to the conversation)
What I kept:      (which output files, unchanged)
What I changed:   (and why; "nothing" is the expected answer today)
What I verified:  (which acceptance criteria you ran, and the results)
