Reproduce a customer report and attach the failing test
A customer report turned into a failing test, even when the fix itself needs a product decision.
The ticket
Turn a customer report into a reproducible failing test.
Acceptance criteria
- A test reproduces the reported behaviour
- The conditions needed to reproduce it are documented
- If the correct behaviour is unclear, that question is stated rather than answered
- The report is linked
What lands as proof
A failing test derived from the report, which converts an anecdote into something the team can act on.
Why teams defer it
- Reports arrive without versions, environments, or steps, so reproducing one is detective work.
- Support and engineering keep the information in different systems, so the context gets lost between them.
Questions
- What does the agent actually change?
- The ticket is scoped to one outcome: turn a customer report into a reproducible failing test. Work that serves that outcome is in scope, and anything outside it is left for a separate ticket, so the pull request stays reviewable.
- How do I know the work is done?
- The pull request carries the evidence, not only the diff. Here that means the reproduction lands as a pull request the team can act on, so a reviewer can confirm the result without reproducing the work locally.
- How much oversight does this need?
- None is required before the run starts. Every claim this ticket makes can be checked by machine, so the agent works straight through and a person reviews the finished pull request like any other.