Check whether the outcome improved
Define environment outcomes
Write expected endpoint, credential label, resource marker, allowed operation, and local result for matching test and production cases. Define a safe outcome for crossed or unmarked responses. Keep any real environment checks outside the synthetic fixture clearly bounded.
Exercise configuration boundaries
Run the matrix from a clean process, an explicit override, and a restart. Compare selected base URL, request destination, response marker, and stored record. Confirm a test identifier cannot be silently accepted as live work or vice versa.
Check mutation safeguards
Attempt only safe operations in crossed cases, then inspect that the client blocks, labels, or reviews them. Verify logs expose environment and correlation without secrets. Repeat a nearby job or caller that shares configuration before closing the change.
What to carry forward
Accept when matching configurations reach their intended endpoints, crossed or unknown cases stop safely, and labels survive process changes. Keep production evidence partial when it was not controlled. Record fixtures, configuration source, client revision, and review result.
Technical background: OpenAI documentation.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow