Question 800
At Bluehaven University, an internal student services team is improving student advising following a quarterly access review (SF-P01-01). The administrator observes that a support team copies case comments into a consumer chatbot to create CRM case summaries. Any fix must ensure that student records never cross program authorization boundaries. Which observable result would best demonstrate that the correction actually worked?
Explanation
Correct Answer
This is the direct acceptance signal for the reported failure: In a controlled Student_Request__c acceptance test, an authorized agent invokes the record-field prompt and the summary references the intended case only; the record is retained under Bluehaven University's rollout (SF-P01-01). The chosen correction was to use Prompt Builder with a record-grounded field generation template for the approved case summary field; the measurement therefore tests the relevant source, routing or authorization outcome rather than unrelated activity.
Incorrect Answers
The proposed observation “The cleanup runs under documented business rules without a generative response being used as the control; confirm this while testing the unauthorized case” would check a different original case: A nightly data cleanup requirement is being treated as a request for open-ended generated prose. The correction must ensure that customer order details stay within each authorized account. Its mechanism is: Prompt Builder generates contextual text and should not replace deterministic business-rule execution. In the present case, a support team copies case comments into a consumer chatbot to create CRM case summaries. The correction must ensure that student records never cross program authorization boundaries; the more direct acceptance evidence is “An authorized agent invokes the record-field prompt and the summary references the intended case only; confirm this using a least-privilege test identity” because prompt Builder creates reusable Salesforce-grounded prompt templates instead of unmanaged external text entry.
The proposed observation “Two eligible users generate consistent product descriptions from the same approved source fields; confirm this and compare with approved baseline behavior” would check a different original case: Regional teams handwrite inconsistent product descriptions despite having approved CRM product facts. The correction must ensure that supplier prices are never shown outside entitled teams. Its mechanism is: A reusable prompt template standardizes instructions and connects them to trusted contextual data. In the present case, a support team copies case comments into a consumer chatbot to create CRM case summaries. The correction must ensure that student records never cross program authorization boundaries; the more direct acceptance evidence is “An authorized agent invokes the record-field prompt and the summary references the intended case only; confirm this using a least-privilege test identity” because prompt Builder creates reusable Salesforce-grounded prompt templates instead of unmanaged external text entry.
The proposed observation “A malicious-article test does not trigger unauthorized actions or policy overrides; confirm this while testing the unauthorized case” would check a different original case: A retrieved knowledge article includes instructions telling the assistant to ignore the company policy. The correction must ensure that work order updates follow supported automation guardrails. Its mechanism is: External content must not become trusted operational instruction merely because it is retrieved. In the present case, a support team copies case comments into a consumer chatbot to create CRM case summaries. The correction must ensure that student records never cross program authorization boundaries; the more direct acceptance evidence is “An authorized agent invokes the record-field prompt and the summary references the intended case only; confirm this using a least-privilege test identity” because prompt Builder creates reusable Salesforce-grounded prompt templates instead of unmanaged external text entry.