Design business-ready prompts with full control and clarity
Creating prompts is at the core of using AI effectively in Salesforce. GPTfy offers a purpose-built Prompt Builder that simplifies prompt creation, testing, and deployment—all in one structured interface.
With intuitive controls, minimal clicks, and smart defaults, teams can quickly create, test, and deploy business-specific prompts that align with both operational needs and AI best practices.
Before You Start
- Make sure the following prerequisites are completed:
- Activate your preferred AI model
- Data Context Mapping should be configured.
Step 1: Create a new prompt
To begin, navigate to the Prompt Builder tab and select New Prompt. Provide the required fields:
- Prompt Name: Enter a clear, descriptive title.
- Target Object: Choose the Salesforce object (e.g., Account, Opportunity).
- Data Context Mapping: Select the appropriate mapping configuration.
- AI Model: Choose the AI model (e.g., OpenAI, Claude etc.).
- Prompt Type:
- Text Prompt – Use your own custom-written prompt command.
- JSON Prompt – Use structured prompt components for advanced workflows.
- Text Prompt – Use your own custom-written prompt command.
- Once filled, click Save. A unique Prompt Request ID will be generated.
Key Fields section (pre-filled)
- Object
- Type: Text / JSON
- Prompt Version
- AI Model
- Status: Draft / Active
- Last Modified
Step 2: Configure the prompt
The Configuration tab contains all foundational settings to define the prompt.
Prompt Configuration section
AI Prompt Name (pre-filled, but editable)
Description: Provide a concise summary of the prompt’s purpose.
Prompt Request ID: Unique backend identifier.
Allow User Input: Enable users to add manual inputs at runtime.
Include Files: Allow file attachments to be processed with the prompt. If enabled, we can select a file type AI Model to process various types of files including- pdf, ppt, docs, xlsx, png, jpg, and jpeg.
AI Model – File Processing: For this field, the user has to create an AI Model of the file type to use the model for file processing. Then select the same AI model to here to perform this functionality.
Purpose: Multi-select business purposes.
How It Works: Rich-text explanation of the prompt’s logic and behavior.
Prompt Availability Settings section
Visibility Condition: Add filter logic to control where prompts appear
Visibility Exception Message: Custom message when prompt can’t be run
Profile Access: Limit prompt access by Salesforce user profile
Note: Any errors during prompt execution or validation now appear as a clear banner at the top of the page for better visibility.
AI Model section
Specify AI-related parameters:
AI Model: The engine powering the prompt
Temperature: Controls response creativity (0–1)
Top P: Influences the diversity of considered responses
Max Output Tokens: Sets the limit on the AI’s response length
System Information section
View details such as Prompt Version, Message, Owner ID, Created By, Created Date, Last Modified By, and Last Modified Date.
Quick Tip: You can now press Ctrl+S (or Cmd+S on Mac) to quickly save your prompt while working in the builder.
Step 3: Set Grounding Rules and Locale
The Grounding tab ensures ethical, brand-safe AI behavior.
Grounding Rules section
Select from predefined ethical and content guidelines, such as:
- Ethical Grounding Examples:
- Prioritize safety and ethics.
- Avoid using scare tactics.
- Base responses solely on provided prompt information.
- Reference linked JSON records correctly.
- Context Grounding Examples:
- Only answer what is asked.
- Refrain from using first-person pronouns.
- Ensure all related records are shown accurately.
Locale Configuration section
Ensures prompts respond based on user or automation locale
| Use Case | Option | Description |
|---|---|---|
|
Automation |
Company Default |
Uses standard org-wide locale
|
|
|
Record Owner |
Uses locale of record owner
|
|
GPTfy Console Use |
Active User |
Uses logged-in user’s locale |
|
|
Company Default |
Maintains uniform experience across users |
Step 4: Define the Prompt Command
The Prompt tab allows you to write or assemble the AI instruction.
For Text Prompts
- Use the Prompt Command text box to enter free-form instructions.
- Update: The Prompt Command Box Size has been increased to make writing longer and more complex prompts easier.
- Insert Merge Fields to dynamically pull Salesforce field values.
For JSON Prompts
- Define modular sub-components for structured prompting.
You can also:
- Record Lookup: Search and select records from the target object.
- New Test Tab: Now you can test prompts under a dedicated Test Tab, where the value in the select record field is automatically populated with the previous test runs based on the security audit records.
- Quick Test: Instantly validate prompts on selected records before full activation.
Each test execution logs detailed inputs, outputs, and timestamps for traceability.
Step 5: Add Post-Prompt Actions
The Actions tab defines automation triggered after a prompt runs.
Supported Post-Prompt Actions
- Update Field: Modify a field value on the current record.
- Create Record: Create a new Salesforce record.
- Invoke Flow: Trigger an autolaunched Salesforce Flow.
- Invoke Apex: Execute an Apex class.
- Open Form: Display a pre-filled form based on AI output.
Each action can
- Use hardcoded, field-reference, or AI-generated values.
- Be sequenced for complex workflows.
Enhancement
Prompt Action Validation ensures that only active auto-launched flows and valid Apex classes can be selected, avoiding configuration errors.
Step 6: Save, Activate, and Manage
- Click Save to register the prompt
- After building your prompt:
- Save: Store your progress.
- Activate: Publish the prompt for end-user access (must pass validations).
Upon Activation
- The prompt is stored as a JSON file under Salesforce Files.
- It becomes visible to allowed profiles on the target object’s record page.
Editing Prompts
- Deactivate the prompt (moves to Draft status) before editing.
- Editable fields will become active once deactivated.
- Use Clone with Related to duplicate prompts along with their associated mappings and configurations.
Step 7: Version Control, Testing, and Security Audits
GPTfy provides full lifecycle management for prompt governance:
Test Prompt
- Test prompts on real data with live Salesforce records.
- Receive instant validation feedback without needing full deployment.
Version Control
- Every activation saves a new version automatically.
- Use the Reinstate button to roll back to a previous version if needed.
Security Audits
- Every prompt execution is logged with record ID, user, timestamp, and inputs/outputs.
- View execution history under the Audit tab for compliance tracking.
With GPTfy’s Prompt Builder, you can build, test, and deploy powerful prompts with speed and confidence, without relying on multiple tools or developer hand-holding. The intuitive design ensures alignment between intent and results while giving you full control over visibility, grounding, and AI behavior.
Specify conditions by using query builder; the field Condition Visibility auto-populates.
Visibility Exception Message: This is a text field that Prompt Builder can use to explain why the user cannot run the prompt on that particular account. The same message will appear on the GPTfy console when the user hovers over the Run GPTfy Button.
Profile: This multi-select picklist allows the user to decide which profile can access the prompt.
If the user wants to run on the system admin profile The prompt fails to appear in the “Select Prompt” dropdown on the Record page. It becomes visible only when the user is on the correct profile.same as given on prompt configuration.
Profile – This is a multi-select picklist by which the user can decide which profile user can access the prompt.
AI Model: This section contains the details regarding the AI Model settings like Name, temp, Top P, Max Output Tokens
- AI Model: The user can add the AI Model that will power the prompt (e.g., GPTfy, OpenAI).
- Temperature: This field Controls the balance between creativity and predictability in the output (0-1). Higher values lead to more diverse responses.
- Top P: This field decides how many possible words the AI considers, influencing the response’s diversity. Higher values lead to more unique results.
- Max Output Tokens: This field allows the user to set the maximum length of the generated response.
Step 3: Grounding (Ethical & Content)
Grounding—This tab allows you to add “extra flavor” to your prompt command, making the response more tailored to your specific needs.
- Examples might include avoiding bias, promoting fairness, or staying on topic.
Sections:
- Grounding rules: These are multi-select picklist values. When selected and saved, they are appended to the prompt command during activation, influencing the AI’s generation process.
Locale Configuration in GPTfy
GPTfy derives the locale dynamically based on how the prompt is called. This locale information is appended to the prompt command and sent to the AI, ensuring personalized and accurate responses based on context. Below are the details for configuring locale settings:
Locale Options and Usage
Locale – Automation
- Description: Used when the prompt is executed through AI Mass Processing, Flows, or REST API.
- Options:
- Company Default: Applies the standard organization-wide locale for consistent automated formatting.
- Record Owner: Adopts the locale of the record owner to personalize date, time, and currency formatting.
Locale – GPTfy Console
- Description: Used when the prompt is executed from the GPTfy User Console.
- Options:
- Active User: Uses the current user’s locale for a tailored experience.
- Company Default: Applies the standard organization-wide locale for consistency across operations.
Step 4: Prompt Command and Quick testing
Prompt – This tab shows the Prompt command and enables users to search the record and run the prompt.
- Prompt Command—A prompt Command is a text box where the user can enter the command as expected. If the prompt is a text prompt, a prompt command text box will appear, and if the prompt type is JSON, the prompt component list will appear.
Users can search the target object’s record from the lookup below and run the GPTfy for quick testing, even if the prompt is inactive. A security audit record will be created after execution.
Step 5: Actions
The Prompt Action Framework is a functionality within the Prompt Builder that allows users to perform various actions on a record when a prompt is run. These actions are specific to each prompt and can be configured in the Actions Tab of the prompt setup. For more details, Click Here.
Step 6: Save & Activate
Click “Save” to create your prompt. Once saved, a unique “Prompt Request ID” appears for reference.
Click “Activate” to make it available in the GPTfy Console dropdown. This will trigger a validation check. Upon successful validation, your prompt is ready to use!
Editing the Prompt—There are no separate buttons for editing the prompt. When the user deactivates the prompt, all prompt fields become editable. Upon deactivation, the prompt’s status is set to “Draft.”
The user can then Activate the prompt again and make a clone of it by using the ‘Clone with Related’ button.
Step 7: Version Control and Audits
- Enable the prompt versioning from the preferences tab if the AI setting is enabled to enable the functionality.
Version Control: Every activation creates a JSON version that is stored under Files for tracking. Track changes and revert to previous versions if needed.
- The “Reinstate” button allows you to restore an older version.
Security Audit: The user can check the security audit records beside the Versions tab. Whenever the user runs the prompt, the security audit record is added.
