Prompt Engineering20% of the exam
Prompt Builder use cases and solution fit
When a requirement calls for a Prompt Builder template (generated language grounded in Salesforce data that admins can tune) and when it calls for a formula, roll-up summary, flow or validation rule because the result must be exact, repeatable and auditable.
Given business requirements, identify when it's appropriate to use Prompt Builder.
~8 min read · 7 practice questions ·
What the exam expects
- Decide from a business requirement whether the deliverable is generated language (a Prompt Builder job) or an exact, repeatable value or rule (a formula, roll-up summary, flow or validation rule).
- Recognise the signals that favour Prompt Builder: natural-language output, grounded in record data, reused across many records, and tuned by admins without code.
- Explain why a requirement for identical output on identical inputs, or for an auditable figure, is not met by any prompt template configuration.
- Match where the output must appear (record field, email composer, flow, Apex, API or agent action) to the template that serves it.
- Key terms: prompt template (a reusable prompt with placeholders that Prompt Builder fills with your data); grounding (adding Salesforce data to a prompt for context); prompt resolution (replacing the placeholders with real values before the prompt goes to the LLM); deterministic (same inputs, same output, every time); entry point (where users or automation call the template).
Key facts
What Prompt Builder manages
- Detail
- Reusable prompt templates grounded with merge fields that reference record fields, flows, Apex and uploaded files
Repeatability
- Detail
- Every run is a new LLM call; responses can differ for the exact same prompt
Licensing
- Detail
- Enterprise, Performance or Unlimited with Einstein for Platform, an Einstein or Agentforce for Sales or Service add-on, or Agentforce Foundations
Run-time data access
- Detail
- Merge-field data is governed by the executing user's permissions
Trust Layer scope
- Detail
- Applies only to generative AI and Agentforce features
Field generation target
- Detail
- Writable text, text area, text area (long) and text area (rich); not read-only fields
Flex entry points
- Detail
- None predefined; Apex, Flow, REST API or Agentforce
Calling any template
- Detail
- Invocable action, Connect REST API or Connect in Apex
Roll-up summary
- Detail
- Count, sum, min or max of detail records over a master-detail relationship
Sources for this section (8)
- Prompt Builder Key Conceptshelp.salesforce.com (opens in a new tab)
- Create a Prompt Templatehelp.salesforce.com (opens in a new tab)
- Run-time Executionhelp.salesforce.com (opens in a new tab)
- Einstein Trust Layer: Designed for Trusthelp.salesforce.com (opens in a new tab)
- Build a Field Generation Prompt Templatetrailhead.salesforce.com (opens in a new tab)
- Create a Flex Prompt Templatetrailhead.salesforce.com (opens in a new tab)
- Integrate Prompt Templateshelp.salesforce.com (opens in a new tab)
- Roll-Up Summary Fieldhelp.salesforce.com (opens in a new tab)
- Why can no prompt template configuration guarantee identical output for identical inputs?
- When should a requirement get a flex template instead of a field generation template?
- When is a roll-up summary field the better answer than a prompt template?
- Which Salesforce resources can ground a Prompt Builder template?
The questions are free to read. Answers with their sources and self-grading are included with the pack.
Keep reading with the Agentforce Specialist pack
- Core concepts
- Exam traps
- Worked scenario
The pack includes:
- All 26 lessons of the Agentforce Specialist guide, with reading progress
- Interactive decision walks and step-by-step sequence steppers
- Inline practice questions and one-click topic practice (7 questions for this lesson)
- An official source cited for every claim
- An AI tutor that answers from the guide
- Unlimited timed practice exams