A good MR description helps a reviewer understand the reason for a change and decide how to check it. The diff alone often can't tell Duo why the change was made or which checks you've actually run.
Supply the diff and the context
Use this diff in the original prompt:
def shipping_fee(subtotal):
- return 0 if subtotal > 50 else 5
+ return 0 if subtotal >= 50 else 5
Add the following fictional MR context beneath it:
The agreed shipping rule is free shipping for orders of $50 or more.
Orders worth exactly $50 were being charged $5.
I ran the three shipping unit tests for subtotals of $49, $50, and $51; all passed.
I haven't run the full suite or checked the checkout page manually.
There is no related issue. Don't invent an issue reference.
Review the proposed description
The following illustrative draft shows how the supplied context could become an MR description. You can compare your answer with it:
Orders worth exactly $50 were still being charged for shipping. This change includes $50 in the free-shipping threshold, matching the agreed rule.
The three shipping unit tests pass for $49, $50, and $51. The full suite and a manual checkout check haven't been run yet.
To check the change, confirm that shipping costs $5 at $49 and is free at $50 and $51. The function's inputs and return type haven't changed; the fee at exactly $50 has.
The reviewer can see which checks you've completed and which they could still run. Saying “All tests pass” would overstate the supplied results, and adding an issue number would link the fix to work you haven't identified.
Add a result when you have it
You can continue the conversation after testing the checkout page:
I checked the checkout page at $49 and $50. It displayed the expected shipping
fees in both cases. Update the description to include that result, while keeping
the full-suite check marked as not run.
Duo can now include the manual checkout result alongside the unit tests and keep the full suite listed as a check you haven't run.