What you will learn
A creation response is one piece of evidence. A later read of the returned ID tests whether a service actually saved the resource. The legacy users API intentionally simulates changes.
Your goals
- Compare POST success with a later GET.
- Avoid treating a simulated reply as stored data.
Before you start
Read the previous lesson first if you have not already. This course does not require programming.
Practice setup
These requests use the public TestKru teaching API. No workspace is needed unless a step says so. Leave the body empty for GET, HEAD and OPTIONS. Example people and emails are fictional. Changes to /api/users are simulated and are not saved.
Follow these steps
- Send the users POST and select its response. Read the returned id, name and email.
- Read the API note that legacy changes are simulated. Send the GET for the id that the create returned. If that id is not 6, say so in your notes instead of inventing a response.
- Select the GET response and compare it with the create response.
- Copy the created id and the read-back status. Choose whether this API saved the user.
- Submit Answers. The later saved-customer lesson is where a read really finds the created customer.
Investigate before concluding
First record the actual result for each case. Compare the body, headers and later state with the stated contract. Keep an observation separate from a prediction; use the model answer only after making your own attempt.
What to keep as evidence
- The returned ID is used in the follow-up.
- Both responses are recorded.
- Simulation is distinguished from a violated persistence contract.
References for this lesson
Finish this assignment
Select each named response with Use this response. Answer from those responses or the transcript on this page. Click Submit Answers. Your score is the percentage of checks that passed. Core checks are required to finish. Extra evidence checks are labelled. A saved 100/100 means every check passed. Continue opens the next lesson you can submit.
If something goes wrong
- Change one part of the request at a time. If a check needs attention, read the note under it and compare it with the response you selected.
Self-review and independent variation
- The returned ID is used in the follow-up.
- Both responses are recorded.
- Simulation is distinguished from a violated persistence contract.
Repeat the POST with a different fictional name and compare the returned ID; do not assume two independent saved users.
Automatic checks verify the requests and answers configured for this lesson. The self-review is for your own reasoning. A score is not a certificate.