├── images ├── AI_Gateway_for_Salesforce_Architecture.png └── SalesforceDevops.net_An_icon_that_is_robot_with_a_wrench.png ├── force-app └── main │ └── default │ ├── contentassets │ ├── SalesforceDevopsnet_An_icon_that_is.asset │ └── SalesforceDevopsnet_An_icon_that_is.asset-meta.xml │ ├── classes │ ├── OpenAIConnect.cls-meta.xml │ ├── OpenAIConnectTest.cls-meta.xml │ ├── OpenAIHttpCalloutMock.cls-meta.xml │ ├── PromptStatsController.cls-meta.xml │ ├── PromptStatsControllerTest.cls-meta.xml │ ├── OpenAIHttpCalloutMock.cls │ ├── PromptStatsController.cls │ ├── PromptStatsControllerTest.cls │ ├── OpenAIConnectTest.cls │ └── OpenAIConnect.cls │ ├── pages │ ├── PromptCoverPage.page-meta.xml │ └── PromptCoverPage.page │ ├── tabs │ ├── Prompt__c.tab-meta.xml │ └── Prompt_Engineering.tab-meta.xml │ ├── objects │ ├── Prompt__c │ │ ├── listViews │ │ │ └── All.listView-meta.xml │ │ ├── fields │ │ │ ├── Tags__c.field-meta.xml │ │ │ ├── Parameters__c.field-meta.xml │ │ │ ├── Title__c.field-meta.xml │ │ │ ├── System__c.field-meta.xml │ │ │ ├── Order__c.field-meta.xml │ │ │ ├── MaxTokens__c.field-meta.xml │ │ │ ├── Prompt__c.field-meta.xml │ │ │ ├── UsedCount__c.field-meta.xml │ │ │ ├── Temperature__c.field-meta.xml │ │ │ ├── PromptCategory__c.field-meta.xml │ │ │ └── Model__c.field-meta.xml │ │ └── Prompt__c.object-meta.xml │ └── PromptAnswer__c │ │ ├── fields │ │ ├── Answer__c.field-meta.xml │ │ ├── PromptUsed__c.field-meta.xml │ │ ├── ParameterValues__c.field-meta.xml │ │ ├── SystemUsed__c.field-meta.xml │ │ ├── Temperature__c.field-meta.xml │ │ ├── TokensUsed__c.field-meta.xml │ │ ├── Prompt__c.field-meta.xml │ │ └── Type__c.field-meta.xml │ │ └── PromptAnswer__c.object-meta.xml │ ├── flexipages │ ├── Prompt_Engineering_UtilityBar.flexipage-meta.xml │ ├── Prompt_Answer_Record_Page.flexipage-meta.xml │ └── Prompt_Record_Page.flexipage-meta.xml │ ├── namedCredentials │ └── OpenAIChat.namedCredential-meta.xml │ ├── applications │ └── PromptEngineering.app-meta.xml │ ├── externalCredentials │ └── OpenAI.externalCredential-meta.xml │ ├── appMenus │ └── AppSwitcher.appMenu-meta.xml │ ├── layouts │ ├── PromptAnswer__c-Prompt Response Layout.layout-meta.xml │ └── Prompt__c-Prompt Layout.layout-meta.xml │ ├── permissionsets │ └── PromptEngineering.permissionset-meta.xml │ ├── flows │ └── PromptProcessing.flow-meta.xml │ └── profiles │ └── Admin.profile-meta.xml ├── scripts ├── open.sh ├── pull.sh ├── push.sh ├── deploy.sh ├── soql │ └── account.soql ├── .env.example ├── apex │ └── hello.apex ├── build.sh └── export.sh ├── .vscode ├── settings.json ├── extensions.json └── launch.json ├── .prettierignore ├── sfdx-project.json ├── .prettierrc ├── .eslintignore ├── Makefile ├── config └── project-scratch-def.json ├── .forceignore ├── .gitignore ├── LICENSE ├── manifest └── package.xml ├── package.json ├── README.md └── data └── Prompt__c.csv /images/AI_Gateway_for_Salesforce_Architecture.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vkeenan/ai-gateway/HEAD/images/AI_Gateway_for_Salesforce_Architecture.png -------------------------------------------------------------------------------- /images/SalesforceDevops.net_An_icon_that_is_robot_with_a_wrench.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vkeenan/ai-gateway/HEAD/images/SalesforceDevops.net_An_icon_that_is_robot_with_a_wrench.png -------------------------------------------------------------------------------- /force-app/main/default/contentassets/SalesforceDevopsnet_An_icon_that_is.asset: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vkeenan/ai-gateway/HEAD/force-app/main/default/contentassets/SalesforceDevopsnet_An_icon_that_is.asset -------------------------------------------------------------------------------- /scripts/open.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | 7 | # Open the scratch org 8 | sfdx org open --target-org $ORG_ALIAS -------------------------------------------------------------------------------- /.vscode/settings.json: -------------------------------------------------------------------------------- 1 | { 2 | "search.exclude": { 3 | "**/node_modules": true, 4 | "**/bower_components": true, 5 | "**/.sfdx": true 6 | }, 7 | "salesforcedx-vscode-core.show-cli-success-msg": false 8 | } 9 | -------------------------------------------------------------------------------- /scripts/pull.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | 7 | # Open the scratch org 8 | sfdx project retrieve start --target-org $ORG_ALIAS -------------------------------------------------------------------------------- /scripts/push.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | 7 | # Open the scratch org 8 | sfdx project deploy start --target-org $ORG_ALIAS -------------------------------------------------------------------------------- /scripts/deploy.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | # 7 | sfdx force:source:deploy --manifest ./manifest/package.xml --target-org $PACKAGE_ALIAS -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIConnect.cls-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | Active 5 | -------------------------------------------------------------------------------- /.vscode/extensions.json: -------------------------------------------------------------------------------- 1 | { 2 | "recommendations": [ 3 | "salesforce.salesforcedx-vscode", 4 | "redhat.vscode-xml", 5 | "dbaeumer.vscode-eslint", 6 | "esbenp.prettier-vscode", 7 | "financialforce.lana" 8 | ] 9 | } 10 | -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIConnectTest.cls-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | Active 5 | -------------------------------------------------------------------------------- /force-app/main/default/pages/PromptCoverPage.page-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | 5 | -------------------------------------------------------------------------------- /.prettierignore: -------------------------------------------------------------------------------- 1 | # List files or directories below to ignore them when running prettier 2 | # More information: https://prettier.io/docs/en/ignore.html 3 | # 4 | 5 | **/staticresources/** 6 | .localdevserver 7 | .sfdx 8 | .vscode 9 | 10 | coverage/ -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIHttpCalloutMock.cls-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | Active 5 | -------------------------------------------------------------------------------- /force-app/main/default/classes/PromptStatsController.cls-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | Active 5 | -------------------------------------------------------------------------------- /force-app/main/default/classes/PromptStatsControllerTest.cls-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 57.0 4 | Active 5 | -------------------------------------------------------------------------------- /force-app/main/default/tabs/Prompt__c.tab-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | true 4 | Custom32: Factory 5 | 6 | -------------------------------------------------------------------------------- /scripts/soql/account.soql: -------------------------------------------------------------------------------- 1 | // Use .soql files to store SOQL queries. 2 | // You can execute queries in VS Code by selecting the 3 | // query text and running the command: 4 | // SFDX: Execute SOQL Query with Currently Selected Text 5 | 6 | SELECT Id, Name FROM Account 7 | -------------------------------------------------------------------------------- /sfdx-project.json: -------------------------------------------------------------------------------- 1 | { 2 | "packageDirectories": [ 3 | { 4 | "path": "force-app", 5 | "default": true 6 | } 7 | ], 8 | "name": "sf-prompts", 9 | "namespace": "", 10 | "sfdcLoginUrl": "https://login.salesforce.com", 11 | "sourceApiVersion": "58.0" 12 | } 13 | -------------------------------------------------------------------------------- /force-app/main/default/tabs/Prompt_Engineering.tab-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | Custom64: Compass 5 | PromptCoverPage 6 | 7 | -------------------------------------------------------------------------------- /.prettierrc: -------------------------------------------------------------------------------- 1 | { 2 | "trailingComma": "none", 3 | "overrides": [ 4 | { 5 | "files": "**/lwc/**/*.html", 6 | "options": { "parser": "lwc" } 7 | }, 8 | { 9 | "files": "*.{cmp,page,component}", 10 | "options": { "parser": "html" } 11 | } 12 | ] 13 | } 14 | -------------------------------------------------------------------------------- /.eslintignore: -------------------------------------------------------------------------------- 1 | **/lwc/**/*.css 2 | **/lwc/**/*.html 3 | **/lwc/**/*.json 4 | **/lwc/**/*.svg 5 | **/lwc/**/*.xml 6 | **/aura/**/*.auradoc 7 | **/aura/**/*.cmp 8 | **/aura/**/*.css 9 | **/aura/**/*.design 10 | **/aura/**/*.evt 11 | **/aura/**/*.json 12 | **/aura/**/*.svg 13 | **/aura/**/*.tokens 14 | **/aura/**/*.xml 15 | **/aura/**/*.app 16 | .sfdx 17 | -------------------------------------------------------------------------------- /Makefile: -------------------------------------------------------------------------------- 1 | .PHONY: build open push pull export sync deploy 2 | 3 | build: 4 | ./scripts/build.sh 5 | 6 | open: 7 | ./scripts/open.sh 8 | 9 | push: 10 | ./scripts/push.sh 11 | 12 | pull: 13 | ./scripts/pull.sh 14 | 15 | export: 16 | ./scripts/export.sh 17 | 18 | sync: 19 | ./scripts/push.sh 20 | ./scripts/pull.sh 21 | 22 | deploy: 23 | ./scripts/deploy.sh -------------------------------------------------------------------------------- /config/project-scratch-def.json: -------------------------------------------------------------------------------- 1 | { 2 | "orgName": "Prompt Engineering", 3 | "edition": "Developer", 4 | "features": ["EnableSetPasswordInApi"], 5 | "settings": { 6 | "lightningExperienceSettings": { 7 | "enableS1DesktopEnabled": true 8 | }, 9 | "mobileSettings": { 10 | "enableS1EncryptedStoragePref2": false 11 | } 12 | } 13 | } 14 | -------------------------------------------------------------------------------- /scripts/.env.example: -------------------------------------------------------------------------------- 1 | export ORG_ALIAS="ai-gateway-scratch" # replace with your preferred scratch org alias 2 | export PACKAGE_ALIAS="ai-gateway-package" # user alias for packaging org 3 | export CONFIG_FILE="config/project-scratch-def.json" # replace with path to your scratch org definition file if different 4 | export DEV_HUB_ALIAS="you@yourdomain.com" # replace with your Dev Hub alias or username 5 | -------------------------------------------------------------------------------- /.forceignore: -------------------------------------------------------------------------------- 1 | # List files or directories below to ignore them when running force:source:push, force:source:pull, and force:source:status 2 | # More information: https://developer.salesforce.com/docs/atlas.en-us.sfdx_dev.meta/sfdx_dev/sfdx_dev_exclude_source.htm 3 | # 4 | 5 | package.xml 6 | 7 | # LWC configuration files 8 | **/jsconfig.json 9 | **/.eslintrc.json 10 | 11 | # LWC Jest 12 | **/__tests__/** -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/listViews/All.listView-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | All 4 | NAME 5 | Title__c 6 | UsedCount__c 7 | Model__c 8 | Everything 9 | 10 | 11 | -------------------------------------------------------------------------------- /scripts/apex/hello.apex: -------------------------------------------------------------------------------- 1 | // Use .apex files to store anonymous Apex. 2 | // You can execute anonymous Apex in VS Code by selecting the 3 | // apex text and running the command: 4 | // SFDX: Execute Anonymous Apex with Currently Selected Text 5 | // You can also execute the entire file by running the command: 6 | // SFDX: Execute Anonymous Apex with Editor Contents 7 | 8 | string tempvar = 'Enter_your_name_here'; 9 | System.debug('Hello World!'); 10 | System.debug('My name is ' + tempvar); -------------------------------------------------------------------------------- /.vscode/launch.json: -------------------------------------------------------------------------------- 1 | { 2 | // Use IntelliSense to learn about possible attributes. 3 | // Hover to view descriptions of existing attributes. 4 | // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387 5 | "version": "0.2.0", 6 | "configurations": [ 7 | { 8 | "name": "Launch Apex Replay Debugger", 9 | "type": "apex-replay", 10 | "request": "launch", 11 | "logFile": "${command:AskForLogFileName}", 12 | "stopOnEntry": true, 13 | "trace": true 14 | } 15 | ] 16 | } 17 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Tags__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Tags__c 4 | A text list of Tags which identify this Prompt 5 | false 6 | A text list of Tags which identify this Prompt 7 | 8 | false 9 | false 10 | TextArea 11 | 12 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/Answer__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Answer__c 4 | The Response from the Prompt 5 | false 6 | The Response from the Prompt 7 | 8 | 32768 9 | false 10 | LongTextArea 11 | 20 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Parameters__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Parameters__c 4 | JSON parameter definitions 5 | false 6 | JSON parameter definitions 7 | 8 | 2048 9 | false 10 | LongTextArea 11 | 10 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Title__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Title__c 4 | The Title for this Prompt 5 | false 6 | The Title for this Prompt 7 | 8 | 255 9 | false 10 | false 11 | Text 12 | false 13 | 14 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/PromptUsed__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | PromptUsed__c 4 | The text of the Prompt message used to generate the Answer text 5 | false 6 | The text of the Prompt message used to generate the Answer text 7 | 8 | 8196 9 | false 10 | LongTextArea 11 | 3 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/ParameterValues__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | ParameterValues__c 4 | JSON body of Parameter Values that generated this Response 5 | false 6 | JSON body of Parameter Values that generated this Response 7 | 8 | 2048 9 | false 10 | LongTextArea 11 | 10 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/SystemUsed__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | SystemUsed__c 4 | The System Message used in this prompt that generated the Answer 5 | false 6 | The System Message used in this prompt that generated the Answer 7 | 8 | 8196 9 | false 10 | LongTextArea 11 | 3 12 | 13 | -------------------------------------------------------------------------------- /scripts/build.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | # Create a new scratch org 7 | echo "Creating scratch org...stand by..." 8 | sfdx force:org:create -s -f $CONFIG_FILE -a $ORG_ALIAS --target-dev-hub $DEV_HUB_ALIAS 9 | 10 | # Push metadata to the scratch org 11 | sfdx project deploy start --target-org $ORG_ALIAS 12 | sfdx force:user:permset:assign --perm-set-name PromptEngineering --target-org $ORG_ALIAS 13 | # Import sample data 14 | sfdx data import tree \ 15 | -f ./data/Prompt__c.json \ 16 | --target-org $ORG_ALIAS 17 | # Open the scratch org 18 | sfdx org open --target-org $ORG_ALIAS 19 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/Temperature__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Temperature__c 4 | "Creativity Temperature" of the generator 5 | false 6 | "Creativity Temperature" of the generator 7 | 8 | 18 9 | false 10 | 8 11 | false 12 | Number 13 | false 14 | 15 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/TokensUsed__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | TokensUsed__c 4 | The number of LLM Tokens used to generate the Answer 5 | false 6 | The number of LLM Tokens used to generate the Answer 7 | 8 | 10 9 | false 10 | 0 11 | false 12 | Number 13 | false 14 | 15 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/System__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | System__c 4 | The System prompt for Chat Completions or the Instructions prompt for the Edit endpoint 5 | false 6 | The System prompt for Chat Completions or the Instructions prompt for the Edit endpoint 7 | 8 | 32768 9 | false 10 | LongTextArea 11 | 10 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Order__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Order__c 4 | 0 5 | The Order in which a child prompt will be processed 6 | false 7 | The Order in which a child prompt will be processed 8 | 9 | 10 10 | false 11 | 0 12 | false 13 | Number 14 | false 15 | 16 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/MaxTokens__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | MaxTokens__c 4 | The maximum number of tokens to spend on generating a Prompt Answer 5 | false 6 | The maximum number of tokens to spend on generating a Prompt Answer 7 | 8 | 10 9 | false 10 | 0 11 | false 12 | Number 13 | false 14 | 15 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Prompt__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Prompt__c 4 | The text of the Prompt for Completions, the Input field for Edits, and the User message for Chats 5 | false 6 | The text of the Prompt for Completions, the Input field for Edits, and the User message for Chats 7 | 8 | 8196 9 | false 10 | LongTextArea 11 | 10 12 | 13 | -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIHttpCalloutMock.cls: -------------------------------------------------------------------------------- 1 | @isTest 2 | global class OpenAIHttpCalloutMock implements HttpCalloutMock { 3 | // Implement this interface method 4 | global HTTPResponse respond(HTTPRequest req) { 5 | // Create a fake response 6 | HttpResponse res = new HttpResponse(); 7 | res.setHeader('Content-Type', 'application/json'); 8 | res.setBody('{"id": "1", "object": "object", "created": 1641234123, "model": "gpt-3.5-turbo", "usage": {"prompt_tokens": 10, "completion_tokens": 10, "total_tokens": 20}, "choices": [{"message": {"role": "assistant", "content": "The weather is sunny."}, "finish_reason": "stop", "index": 0}]}'); 9 | res.setStatusCode(200); 10 | return res; 11 | } 12 | 13 | } 14 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/UsedCount__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | UsedCount__c 4 | 0 5 | The count of how many times this prompt has been used 6 | false 7 | The count of how many times this prompt has been used 8 | 9 | 10 10 | false 11 | 0 12 | false 13 | Number 14 | false 15 | 16 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/Prompt__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Prompt__c 4 | SetNull 5 | The Prompt from which this Response was generated 6 | false 7 | The Prompt from which this Response was generated 8 | 9 | Prompt__c 10 | PromptResponses 11 | false 12 | false 13 | Lookup 14 | 15 | -------------------------------------------------------------------------------- /scripts/export.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" 3 | 4 | # Variable Declarations 5 | source $DIR/.env 6 | 7 | sfdx data export tree --output-dir ./data \ 8 | --target-org $ORG_ALIAS \ 9 | --query \ 10 | "select id, name, model__c, order__c, \ 11 | parameters__c, UsedCount__c, \ 12 | tags__c, prompt__c, system__c, title__c, maxtokens__c, \ 13 | temperature__c \ 14 | from Prompt__c" 15 | # 16 | sfdx data query \ 17 | --target-org $ORG_ALIAS \ 18 | --query \ 19 | "select id, name, model__c, order__c, \ 20 | parameters__c, UsedCount__c, \ 21 | tags__c, prompt__c, system__c, title__c, maxtokens__c, \ 22 | temperature__c \ 23 | from Prompt__c" > ./data/Prompt__c.csv 24 | # 25 | -------------------------------------------------------------------------------- /force-app/main/default/flexipages/Prompt_Engineering_UtilityBar.flexipage-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | utilityItems 5 | Region 6 | 7 | 8 | backgroundComponents 9 | Background 10 | 11 | Prompt Engineering UtilityBar 12 | 19 | UtilityBar 20 | 21 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Temperature__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Temperature__c 4 | 0.1 5 | The "Creativity Temperature" to use in generating the Prompt Answer 6 | false 7 | The "Creativity Temperature" to use in generating the Prompt Answer 8 | 9 | 18 10 | false 11 | 8 12 | false 13 | Number 14 | false 15 | 16 | -------------------------------------------------------------------------------- /force-app/main/default/contentassets/SalesforceDevopsnet_An_icon_that_is.asset-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | false 4 | en_US 5 | SalesforceDevopsnet_An_icon_that_is 6 | 7 | 8 | VIEWER 9 | 10 | 11 | 12 | 13 | 1 14 | SalesforceDevops.net_An_icon_that_is_robot_with_a_wrench_in_one_cd929c3e-0e77-46d6-817b-f53f34b73c99.png 15 | 16 | 17 | 18 | -------------------------------------------------------------------------------- /force-app/main/default/classes/PromptStatsController.cls: -------------------------------------------------------------------------------- 1 | /** 2 | * @description This class serves as a controller for retrieving statistics about Prompt records and PromptAnswer records. 3 | * It's generally used to display these statistics on a Visualforce page or a Lightning Component. 4 | * 5 | * The two main properties, `totalPrompts` and `totalAnswers`, represent the total count of records in `Prompt__c` and `PromptAnswer__c` respectively. 6 | */ 7 | 8 | public with sharing class PromptStatsController { 9 | public Integer totalPrompts { 10 | get; set; 11 | } 12 | public Integer totalAnswers { 13 | get; set; 14 | } 15 | 16 | public PromptStatsController() { 17 | totalPrompts = System.Database.countQuery('SELECT count() FROM Prompt__c'); 18 | totalAnswers = System.Database.countQuery('SELECT count() FROM PromptAnswer__c'); 19 | } 20 | 21 | } 22 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # This file is used for Git repositories to specify intentionally untracked files that Git should ignore. 2 | # If you are not using git, you can delete this file. For more information see: https://git-scm.com/docs/gitignore 3 | # For useful gitignore templates see: https://github.com/github/gitignore 4 | 5 | # Salesforce cache 6 | .sf/ 7 | .sfdx/ 8 | .localdevserver/ 9 | deploy-options.json 10 | 11 | # LWC VSCode autocomplete 12 | **/lwc/jsconfig.json 13 | 14 | # LWC Jest coverage reports 15 | coverage/ 16 | 17 | # Logs 18 | logs 19 | *.log 20 | npm-debug.log* 21 | yarn-debug.log* 22 | yarn-error.log* 23 | 24 | # Dependency directories 25 | node_modules/ 26 | 27 | # Eslint cache 28 | .eslintcache 29 | 30 | # MacOS system files 31 | .DS_Store 32 | 33 | # Windows system files 34 | Thumbs.db 35 | ehthumbs.db 36 | [Dd]esktop.ini 37 | $RECYCLE.BIN/ 38 | 39 | # Local environment variables 40 | .env 41 | 42 | # Python Salesforce Functions 43 | **/__pycache__/ 44 | **/.venv/ 45 | **/venv/ 46 | -------------------------------------------------------------------------------- /force-app/main/default/namedCredentials/OpenAIChat.namedCredential-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | false 4 | true 5 | false 6 | 7 | 8 | Url 9 | Url 10 | https://api.openai.com/v1/chat/completions 11 | 12 | 13 | OpenAI 14 | ExternalCredential 15 | Authentication 16 | 17 | SecuredEndpoint 18 | 19 | -------------------------------------------------------------------------------- /force-app/main/default/applications/PromptEngineering.app-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | #EB7E40 5 | SalesforceDevopsnet_An_icon_that_is 6 | 1 7 | true 8 | 9 | Save and Share AI Prompts 10 | Small 11 | Large 12 | false 13 | false 14 | false 15 | 16 | Standard 17 | Prompt_Engineering 18 | Prompt__c 19 | Lightning 20 | Prompt_Engineering_UtilityBar 21 | 22 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2023 Vernon Keenan 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/PromptCategory__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | PromptCategory__c 4 | The base category for this Prompt 5 | false 6 | The base category for this Prompt 7 | 8 | false 9 | false 10 | Picklist 11 | 12 | true 13 | 14 | false 15 | 16 | Devopment 17 | false 18 | 19 | 20 | 21 | Marketing 22 | false 23 | 24 | 25 | 26 | Sales 27 | false 28 | 29 | 30 | 31 | 32 | 33 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/fields/Model__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Model__c 4 | The recommended AI Model for this Prompt 5 | false 6 | The recommended AI Model for this Prompt 7 | 8 | false 9 | false 10 | Picklist 11 | 12 | true 13 | 14 | false 15 | 16 | gpt-3.5-turbo 17 | false 18 | 19 | 20 | 21 | gpt-3.5-turbo-16k 22 | false 23 | 24 | 25 | 26 | gpt-4 27 | false 28 | 29 | 30 | 31 | 32 | 33 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/fields/Type__c.field-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Type__c 4 | "Normal" 5 | The Type of Response 6 | false 7 | The Type of Response 8 | 9 | false 10 | false 11 | Picklist 12 | 13 | true 14 | 15 | false 16 | 17 | Normal 18 | false 19 | 20 | 21 | 22 | Test Result 23 | false 24 | 25 | 26 | 27 | Valid Result 28 | false 29 | 30 | 31 | 32 | 33 | 34 | -------------------------------------------------------------------------------- /manifest/package.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | * 5 | ApexClass 6 | 7 | 8 | * 9 | ApexPage 10 | 11 | 12 | * 13 | CustomApplication 14 | 15 | 16 | * 17 | ContentAsset 18 | 19 | 20 | * 21 | CustomObject 22 | 23 | 24 | * 25 | CustomTab 26 | 27 | 28 | * 29 | ExternalCredential 30 | 31 | 32 | * 33 | FlexiPage 34 | 35 | 36 | * 37 | Flow 38 | 39 | 40 | * 41 | Layout 42 | 43 | 44 | * 45 | NamedCredential 46 | 47 | 48 | * 49 | PermissionSet 50 | 51 | 58.0 52 | 53 | -------------------------------------------------------------------------------- /force-app/main/default/classes/PromptStatsControllerTest.cls: -------------------------------------------------------------------------------- 1 | @isTest 2 | public class PromptStatsControllerTest { 3 | @isTest 4 | static void testPromptStatsController() { 5 | // Prepare test data 6 | // Insert test Prompts 7 | List testPrompts = new List(); 8 | List testPromptAnswers = new List(); 9 | 10 | for (Integer i = 0; i < 5; i++) { 11 | testPrompts.add(new Prompt__c( 12 | // Assume you have Name field 13 | // You should replace this with actual field names and data for your Prompt__c object 14 | Name = 'Test Prompt ' + i 15 | )); 16 | } 17 | 18 | insert testPrompts; 19 | 20 | // Insert test Prompt Answers 21 | 22 | for (Integer i = 0; i < 7; i++) { 23 | testPromptAnswers.add(new PromptAnswer__c( 24 | // Assume you have Name field 25 | // You should replace this with actual field names and data for your PromptAnswer__c object 26 | Prompt__c = testPrompts[0].Id, 27 | Name = 'Test Prompt Answer ' + i 28 | )); 29 | } 30 | 31 | insert testPromptAnswers; 32 | 33 | Test.startTest(); 34 | // Create an instance of the controller 35 | PromptStatsController controller = new PromptStatsController(); 36 | Test.stopTest(); 37 | 38 | // Verify the results 39 | System.assertEquals(5, controller.totalPrompts, 'Total prompts should match the number of inserted prompts'); 40 | System.assertEquals(7, controller.totalAnswers, 'Total answers should match the number of inserted prompt answers'); 41 | } 42 | 43 | } 44 | -------------------------------------------------------------------------------- /package.json: -------------------------------------------------------------------------------- 1 | { 2 | "name": "sf-prompts", 3 | "private": true, 4 | "version": "1.7", 5 | "description": "AI Gateway for Salesforce", 6 | "author": { 7 | "name": "Vernon Keenan", 8 | "url": "https://salesforcedevops.net", 9 | "email": "vern@salesforcedevops.net" 10 | }, 11 | "scripts": { 12 | "lint": "eslint **/{aura,lwc}/**", 13 | "test": "npm run test:unit", 14 | "test:unit": "sfdx-lwc-jest", 15 | "test:unit:watch": "sfdx-lwc-jest --watch", 16 | "test:unit:debug": "sfdx-lwc-jest --debug", 17 | "test:unit:coverage": "sfdx-lwc-jest --coverage", 18 | "prettier": "prettier --write \"**/*.{cls,cmp,component,css,html,js,json,md,page,trigger,xml,yaml,yml}\"", 19 | "prettier:verify": "prettier --list-different \"**/*.{cls,cmp,component,css,html,js,json,md,page,trigger,xml,yaml,yml}\"", 20 | "precommit": "lint-staged" 21 | }, 22 | "devDependencies": { 23 | "@lwc/eslint-plugin-lwc": "^1.1.2", 24 | "@prettier/plugin-xml": "^2.0.1", 25 | "@salesforce/eslint-config-lwc": "^3.2.3", 26 | "@salesforce/eslint-plugin-aura": "^2.0.0", 27 | "@salesforce/eslint-plugin-lightning": "^1.0.0", 28 | "@salesforce/sfdx-lwc-jest": "^1.1.0", 29 | "eslint": "^8.11.0", 30 | "eslint-plugin-import": "^2.25.4", 31 | "eslint-plugin-jest": "^26.1.2", 32 | "lint-staged": "^12.3.7", 33 | "prettier": "^2.6.0", 34 | "prettier-plugin-apex": "^1.10.0" 35 | }, 36 | "lint-staged": { 37 | "**/*.{cls,cmp,component,css,html,js,json,md,page,trigger,xml,yaml,yml}": [ 38 | "prettier --write" 39 | ], 40 | "**/{aura,lwc}/**": [ 41 | "eslint" 42 | ] 43 | } 44 | } 45 | -------------------------------------------------------------------------------- /force-app/main/default/externalCredentials/OpenAI.externalCredential-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | Custom 4 | 5 | PromptEngineering 6 | PromptEngineering 7 | NamedPrincipal 8 | 1 9 | 10 | 11 | DefaultGroup 12 | Content-Type 13 | AuthHeader 14 | application/json 15 | 1 16 | 17 | 18 | DefaultGroup 19 | OpenAI-Organization 20 | AuthHeader 21 | {!$Credential.OpenAI.Org} 22 | 2 23 | 24 | 25 | DefaultGroup 26 | Authorization 27 | AuthHeader 28 | Bearer {!$Credential.OpenAI.apiKey} 29 | 3 30 | 31 | 32 | 33 | -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIConnectTest.cls: -------------------------------------------------------------------------------- 1 | @isTest 2 | public class OpenAIConnectTest { 3 | @isTest static void testConnect() { 4 | // Set up your test data 5 | List inputs = new List(); 6 | OpenAIConnect.openAiInputs testInput = new OpenAIConnect.openAiInputs(); 7 | testInput.promptParameters = '{"placename": "San Francisco"}'; 8 | testInput.openaiSystem = 'You are a helpful assistant.'; 9 | testInput.openaiChatHistory = ''; 10 | testInput.openaiPrompt = 'What is the weather like in {placename} today?'; 11 | testInput.settingsModel = ''; 12 | testInput.settingsTokens = null; 13 | testInput.settingsTemperature = null; 14 | testInput.settingsEndpoint = 'OpenAI_Endpoint'; 15 | inputs.add(testInput); 16 | 17 | // Prepare the test for the HttpRequest 18 | Test.setMock(HttpCalloutMock.class, new OpenAIHttpCalloutMock()); 19 | 20 | // Test the method 21 | Test.startTest(); 22 | List outputs = OpenAIConnect.connect(inputs); 23 | Test.stopTest(); 24 | 25 | // Validate the outputs 26 | System.assertEquals(1, outputs.size(), 'Expected 1 output'); 27 | OpenAIConnect.openAiOutputs output = outputs[0]; 28 | 29 | System.assertEquals('1', output.replyID, 'Expected replyID to match'); 30 | System.assertEquals('object', output.openAiObject, 'Expected object type to match'); 31 | System.assertEquals('gpt-3.5-turbo', output.openAiModel, 'Expected model to match'); 32 | System.assertEquals(20, output.tokensUsed, 'Expected tokens used to match'); 33 | System.assertEquals('stop', output.finishReason, 'Expected finish reason to match'); 34 | System.assertEquals('The weather is sunny.', output.openAiTextOutput, 'Expected first line of chat reply to match'); 35 | } 36 | 37 | } 38 | -------------------------------------------------------------------------------- /force-app/main/default/appMenus/AppSwitcher.appMenu-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | standard__Platform 5 | CustomApplication 6 | 7 | 8 | standard__Sales 9 | CustomApplication 10 | 11 | 12 | standard__Service 13 | CustomApplication 14 | 15 | 16 | standard__Marketing 17 | CustomApplication 18 | 19 | 20 | standard__ServiceConsole 21 | CustomApplication 22 | 23 | 24 | standard__AppLauncher 25 | CustomApplication 26 | 27 | 28 | standard__Community 29 | CustomApplication 30 | 31 | 32 | standard__Sites 33 | CustomApplication 34 | 35 | 36 | standard__Chatter 37 | CustomApplication 38 | 39 | 40 | standard__Content 41 | CustomApplication 42 | 43 | 44 | standard__LightningSales 45 | CustomApplication 46 | 47 | 48 | standard__AllTabSet 49 | CustomApplication 50 | 51 | 52 | CPQIntegrationUserApp 53 | ConnectedApp 54 | 55 | 56 | standard__LightningBolt 57 | CustomApplication 58 | 59 | 60 | standard__SalesforceCMS 61 | CustomApplication 62 | 63 | 64 | PromptEngineering 65 | CustomApplication 66 | 67 | 68 | -------------------------------------------------------------------------------- /force-app/main/default/flexipages/Prompt_Answer_Record_Page.flexipage-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | collapsed 8 | false 9 | 10 | 11 | enableActionsConfiguration 12 | false 13 | 14 | 15 | enableActionsInNative 16 | false 17 | 18 | 19 | hideChatterActions 20 | false 21 | 22 | 23 | numVisibleActions 24 | 3 25 | 26 | force:highlightsPanel 27 | force_highlightsPanel 28 | 29 | 30 | header 31 | Region 32 | 33 | 34 | 35 | 36 | force:detailPanel 37 | force_detailPanel 38 | 39 | 40 | detailTabContent 41 | Facet 42 | 43 | 44 | 45 | 46 | 47 | active 48 | true 49 | 50 | 51 | body 52 | detailTabContent 53 | 54 | 55 | title 56 | Standard.Tab.detail 57 | 58 | flexipage:tab 59 | detailTab 60 | 61 | 62 | maintabs 63 | Facet 64 | 65 | 66 | 67 | 68 | 69 | tabs 70 | maintabs 71 | 72 | flexipage:tabset 73 | flexipage_tabset 74 | 75 | 76 | main 77 | Region 78 | 79 | Prompt Answer Record Page 80 | PromptAnswer__c 81 | 84 | RecordPage 85 | 86 | -------------------------------------------------------------------------------- /force-app/main/default/layouts/PromptAnswer__c-Prompt Response Layout.layout-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | false 5 | false 6 | true 7 | 8 | 9 | 10 | Required 11 | Name 12 | 13 | 14 | Edit 15 | Prompt__c 16 | 17 | 18 | Edit 19 | ParameterValues__c 20 | 21 | 22 | Edit 23 | TokensUsed__c 24 | 25 | 26 | 27 | 28 | Edit 29 | Type__c 30 | 31 | 32 | Edit 33 | OwnerId 34 | 35 | 36 | Edit 37 | Temperature__c 38 | 39 | 40 | 41 | 42 | 43 | true 44 | true 45 | true 46 | 47 | 48 | 49 | Edit 50 | Answer__c 51 | 52 | 53 | Edit 54 | PromptUsed__c 55 | 56 | 57 | Edit 58 | SystemUsed__c 59 | 60 | 61 | 62 | 63 | 64 | false 65 | false 66 | true 67 | 68 | 69 | 70 | Readonly 71 | CreatedById 72 | 73 | 74 | 75 | 76 | Readonly 77 | LastModifiedById 78 | 79 | 80 | 81 | 82 | 83 | true 84 | false 85 | true 86 | 87 | 88 | 89 | 90 | 91 | 92 | false 93 | false 94 | false 95 | false 96 | false 97 | 98 | 00h8D000002lHcp 99 | 4 100 | 0 101 | Default 102 | 103 | 104 | -------------------------------------------------------------------------------- /force-app/main/default/layouts/Prompt__c-Prompt Layout.layout-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | ChangeRecordType 4 | Submit 5 | 6 | false 7 | false 8 | true 9 | 10 | 11 | 12 | Required 13 | Name 14 | 15 | 16 | Edit 17 | Title__c 18 | 19 | 20 | Edit 21 | PromptCategory__c 22 | 23 | 24 | Edit 25 | Model__c 26 | 27 | 28 | Edit 29 | Temperature__c 30 | 31 | 32 | 33 | 34 | Edit 35 | Order__c 36 | 37 | 38 | Edit 39 | Parameters__c 40 | 41 | 42 | Edit 43 | Tags__c 44 | 45 | 46 | Edit 47 | OwnerId 48 | 49 | 50 | Edit 51 | MaxTokens__c 52 | 53 | 54 | 55 | 56 | 57 | true 58 | true 59 | true 60 | 61 | 62 | 63 | Edit 64 | Prompt__c 65 | 66 | 67 | Edit 68 | System__c 69 | 70 | 71 | 72 | 73 | 74 | false 75 | false 76 | true 77 | 78 | 79 | 80 | Readonly 81 | CreatedById 82 | 83 | 84 | 85 | 86 | Readonly 87 | LastModifiedById 88 | 89 | 90 | 91 | 92 | 93 | true 94 | false 95 | true 96 | 97 | 98 | 99 | 100 | 101 | 102 | 103 | NAME 104 | PromptAnswer__c.Prompt__c 105 | 106 | false 107 | false 108 | false 109 | false 110 | false 111 | 112 | 00h7e000002fCtP 113 | 4 114 | 0 115 | Default 116 | 117 | 118 | -------------------------------------------------------------------------------- /force-app/main/default/permissionsets/PromptEngineering.permissionset-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | PromptEngineering 5 | true 6 | 7 | Permissions for the Prompt Engineering project 8 | 9 | true 10 | PromptAnswer__c.Answer__c 11 | true 12 | 13 | 14 | true 15 | PromptAnswer__c.ParameterValues__c 16 | true 17 | 18 | 19 | true 20 | PromptAnswer__c.PromptUsed__c 21 | true 22 | 23 | 24 | true 25 | PromptAnswer__c.Prompt__c 26 | true 27 | 28 | 29 | true 30 | PromptAnswer__c.SystemUsed__c 31 | true 32 | 33 | 34 | true 35 | PromptAnswer__c.Temperature__c 36 | true 37 | 38 | 39 | true 40 | PromptAnswer__c.TokensUsed__c 41 | true 42 | 43 | 44 | true 45 | PromptAnswer__c.Type__c 46 | true 47 | 48 | 49 | true 50 | Prompt__c.MaxTokens__c 51 | true 52 | 53 | 54 | true 55 | Prompt__c.Model__c 56 | true 57 | 58 | 59 | true 60 | Prompt__c.Order__c 61 | true 62 | 63 | 64 | true 65 | Prompt__c.Parameters__c 66 | true 67 | 68 | 69 | true 70 | Prompt__c.PromptCategory__c 71 | true 72 | 73 | 74 | true 75 | Prompt__c.Prompt__c 76 | true 77 | 78 | 79 | true 80 | Prompt__c.System__c 81 | true 82 | 83 | 84 | true 85 | Prompt__c.Tags__c 86 | true 87 | 88 | 89 | true 90 | Prompt__c.Temperature__c 91 | true 92 | 93 | 94 | true 95 | Prompt__c.Title__c 96 | true 97 | 98 | 99 | true 100 | Prompt__c.UsedCount__c 101 | true 102 | 103 | false 104 | 105 | 106 | true 107 | true 108 | true 109 | true 110 | false 111 | PromptAnswer__c 112 | false 113 | 114 | 115 | true 116 | true 117 | true 118 | true 119 | false 120 | Prompt__c 121 | false 122 | 123 | 124 | PromptCoverPage 125 | true 126 | 127 | 128 | Prompt_Engineering 129 | Visible 130 | 131 | 132 | Prompt__c 133 | Visible 134 | 135 | 136 | -------------------------------------------------------------------------------- /force-app/main/default/objects/Prompt__c/Prompt__c.object-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | Accept 5 | Default 6 | 7 | 8 | Accept 9 | Large 10 | Default 11 | 12 | 13 | Accept 14 | Small 15 | Default 16 | 17 | 18 | CancelEdit 19 | Default 20 | 21 | 22 | CancelEdit 23 | Large 24 | Default 25 | 26 | 27 | CancelEdit 28 | Small 29 | Default 30 | 31 | 32 | Clone 33 | Default 34 | 35 | 36 | Clone 37 | Large 38 | Default 39 | 40 | 41 | Clone 42 | Small 43 | Default 44 | 45 | 46 | Delete 47 | Default 48 | 49 | 50 | Delete 51 | Large 52 | Default 53 | 54 | 55 | Delete 56 | Small 57 | Default 58 | 59 | 60 | Edit 61 | Default 62 | 63 | 64 | Edit 65 | Large 66 | Default 67 | 68 | 69 | Edit 70 | Small 71 | Default 72 | 73 | 74 | List 75 | Default 76 | 77 | 78 | List 79 | Large 80 | Default 81 | 82 | 83 | List 84 | Small 85 | Default 86 | 87 | 88 | New 89 | Default 90 | 91 | 92 | New 93 | Large 94 | Default 95 | 96 | 97 | New 98 | Small 99 | Default 100 | 101 | 102 | SaveEdit 103 | Default 104 | 105 | 106 | SaveEdit 107 | Large 108 | Default 109 | 110 | 111 | SaveEdit 112 | Small 113 | Default 114 | 115 | 116 | Tab 117 | Default 118 | 119 | 120 | Tab 121 | Large 122 | Default 123 | 124 | 125 | Tab 126 | Small 127 | Default 128 | 129 | 130 | View 131 | Action override created by Lightning App Builder during activation. 132 | Prompt_Record_Page 133 | Large 134 | false 135 | Flexipage 136 | 137 | 138 | View 139 | Action override created by Lightning App Builder during activation. 140 | Prompt_Record_Page 141 | Small 142 | false 143 | Flexipage 144 | 145 | 146 | View 147 | Default 148 | 149 | false 150 | SYSTEM 151 | Deployed 152 | A Prompt used in a ML orchestration 153 | false 154 | true 155 | false 156 | false 157 | false 158 | true 159 | true 160 | true 161 | true 162 | Private 163 | 164 | 165 | 166 | Text 167 | 168 | Prompts 169 | 170 | ReadWrite 171 | Public 172 | 173 | -------------------------------------------------------------------------------- /force-app/main/default/objects/PromptAnswer__c/PromptAnswer__c.object-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | Accept 5 | Default 6 | 7 | 8 | Accept 9 | Large 10 | Default 11 | 12 | 13 | Accept 14 | Small 15 | Default 16 | 17 | 18 | CancelEdit 19 | Default 20 | 21 | 22 | CancelEdit 23 | Large 24 | Default 25 | 26 | 27 | CancelEdit 28 | Small 29 | Default 30 | 31 | 32 | Clone 33 | Default 34 | 35 | 36 | Clone 37 | Large 38 | Default 39 | 40 | 41 | Clone 42 | Small 43 | Default 44 | 45 | 46 | Delete 47 | Default 48 | 49 | 50 | Delete 51 | Large 52 | Default 53 | 54 | 55 | Delete 56 | Small 57 | Default 58 | 59 | 60 | Edit 61 | Default 62 | 63 | 64 | Edit 65 | Large 66 | Default 67 | 68 | 69 | Edit 70 | Small 71 | Default 72 | 73 | 74 | List 75 | Default 76 | 77 | 78 | List 79 | Large 80 | Default 81 | 82 | 83 | List 84 | Small 85 | Default 86 | 87 | 88 | New 89 | Default 90 | 91 | 92 | New 93 | Large 94 | Default 95 | 96 | 97 | New 98 | Small 99 | Default 100 | 101 | 102 | SaveEdit 103 | Default 104 | 105 | 106 | SaveEdit 107 | Large 108 | Default 109 | 110 | 111 | SaveEdit 112 | Small 113 | Default 114 | 115 | 116 | Tab 117 | Default 118 | 119 | 120 | Tab 121 | Large 122 | Default 123 | 124 | 125 | Tab 126 | Small 127 | Default 128 | 129 | 130 | View 131 | Action override created by Lightning App Builder during activation. 132 | Prompt_Answer_Record_Page 133 | Large 134 | false 135 | Flexipage 136 | 137 | 138 | View 139 | Action override created by Lightning App Builder during activation. 140 | Prompt_Answer_Record_Page 141 | Small 142 | false 143 | Flexipage 144 | 145 | 146 | View 147 | Default 148 | 149 | false 150 | SYSTEM 151 | Deployed 152 | A Response or Answer to a Prompt 153 | false 154 | true 155 | false 156 | false 157 | false 158 | true 159 | true 160 | true 161 | true 162 | Private 163 | 164 | 165 | 166 | Text 167 | 168 | Prompt Answers 169 | 170 | ReadWrite 171 | Public 172 | 173 | -------------------------------------------------------------------------------- /force-app/main/default/flexipages/Prompt_Record_Page.flexipage-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | collapsed 8 | false 9 | 10 | 11 | enableActionsConfiguration 12 | false 13 | 14 | 15 | enableActionsInNative 16 | false 17 | 18 | 19 | hideChatterActions 20 | false 21 | 22 | 23 | numVisibleActions 24 | 3 25 | 26 | force:highlightsPanel 27 | force_highlightsPanel 28 | 29 | 30 | header 31 | Region 32 | 33 | 34 | 35 | 36 | 37 | flowArguments 38 | [{"label":"promptRecordId","dataType":"String","supportsRecordId":true,"value":"{!Record.Id}"}] 39 | 40 | 41 | flowLayout 42 | oneColumn 43 | 44 | 45 | flowName 46 | PromptProcessing 47 | 48 | flowruntime:interview 49 | flowruntime_interview 50 | 51 | 52 | Facet-48daf403-b15f-4057-8c87-eb80f726d10a 53 | Facet 54 | 55 | 56 | 57 | 58 | 59 | relatedListComponentOverride 60 | NONE 61 | 62 | 63 | rowsToDisplay 64 | 10 65 | 66 | 67 | showActionBar 68 | true 69 | 70 | force:relatedListContainer 71 | force_relatedListContainer 72 | 73 | 74 | relatedTabContent 75 | Facet 76 | 77 | 78 | 79 | 80 | force:detailPanel 81 | force_detailPanel 82 | 83 | 84 | detailTabContent 85 | Facet 86 | 87 | 88 | 89 | 90 | 91 | active 92 | true 93 | 94 | 95 | body 96 | Facet-48daf403-b15f-4057-8c87-eb80f726d10a 97 | 98 | 99 | title 100 | OpenAI Prompt 101 | 102 | flexipage:tab 103 | flexipage_tab3 104 | 105 | 106 | 107 | 108 | 109 | body 110 | relatedTabContent 111 | 112 | 113 | title 114 | Answers 115 | 116 | flexipage:tab 117 | relatedListsTab 118 | 119 | 120 | 121 | 122 | 123 | active 124 | false 125 | 126 | 127 | body 128 | detailTabContent 129 | 130 | 131 | title 132 | Standard.Tab.detail 133 | 134 | flexipage:tab 135 | detailTab 136 | 137 | 138 | maintabs 139 | Facet 140 | 141 | 142 | 143 | 144 | 145 | tabs 146 | maintabs 147 | 148 | flexipage:tabset 149 | flexipage_tabset 150 | 151 | 152 | main 153 | Region 154 | 155 | Prompt Record Page 156 | Prompt__c 157 | 160 | RecordPage 161 | 162 | -------------------------------------------------------------------------------- /force-app/main/default/classes/OpenAIConnect.cls: -------------------------------------------------------------------------------- 1 | /** 2 | * OpenAIConnect Class 3 | * 4 | * @description This class provides a connection to the OpenAI API. 5 | * It allows sending prompts to OpenAI, and handles the responses 6 | * received from OpenAI, parsing the data into a usable format for Salesforce. 7 | * 8 | * Main functionality: 9 | * - Prepares the prompts before sending to OpenAI API 10 | * - Sends a HTTP request to OpenAI API with the appropriate prompts and settings 11 | * - Parses the HTTP response from OpenAI API, extracting key information such as 12 | * the generated text, finish reason, response ID and tokens used. 13 | * - Returns a list of outputs containing key information for each input prompt. 14 | * 15 | * Usage: 16 | * An instance of this class is not needed. It exposes a static method "connect" 17 | * which can be called with a list of "openAiInputs" objects. The method returns 18 | * a list of "openAiOutputs" objects, each corresponding to an input prompt. 19 | * 20 | * Example: 21 | * List inputs = new List(); 22 | * List outputs = OpenAIConnect.connect(inputs); 23 | * 24 | * Notes: 25 | * Make sure to handle potential exceptions when calling the "connect" method, 26 | * especially when network issues might occur or when the input prompts are not valid. 27 | * 28 | * Using OpenAIConnect with Salesforce Flow 29 | * 30 | * Salesforce Flows provide a visual way of representing and customizing complex business 31 | * processes. OpenAIConnect can be used within a Flow to facilitate interaction with 32 | * the OpenAI API as part of these processes. 33 | * 34 | * Here's a basic step-by-step guide on how to use OpenAIConnect with a Flow: 35 | * 36 | * 1. Begin by creating a new Flow in Salesforce. In your Salesforce Setup, 37 | * navigate to Flows and click on "New Flow". 38 | * 39 | * 2. Choose the Flow type that suits your needs, generally a Record-Triggered or 40 | * Scheduled-Triggered Flow will be used. 41 | * 42 | * 3. In the Flow Builder, you'll need to create a new Action. This action will 43 | * represent the call to OpenAIConnect.connect method. 44 | * 45 | * 4. Search for and select the "OpenAIConnect" Apex Class and the "connect" method. 46 | * A set of input and output variables corresponding to the "openAiInputs" and 47 | * "openAiOutputs" will be generated. 48 | * 49 | * 5. Set the necessary input fields. These fields can be set to static values, 50 | * or they can be populated dynamically based on other aspects of the Flow. 51 | * 52 | * 6. Save the Action and use the Flow Builder to define what should happen with 53 | * the outputs. For instance, you could create a Record with the OpenAI's 54 | * generated response, or send an Email with its contents. 55 | * 56 | * 7. Complete the Flow by setting the necessary conditions and actions, then 57 | * Save and Activate the Flow. 58 | * 59 | * Note: Be sure to handle potential exceptions and errors in the Flow, as network 60 | * issues might occur or the input prompts might not be valid. 61 | 62 | */ 63 | public class OpenAIConnect { 64 | private static final Decimal DEFAULT_TEMPERATURE = 0.7; 65 | private static final Integer DEFAULT_TOKENS = 1700; 66 | private static final String DEFAULT_SYSTEMPROMPT = 'You are a helpful assistant.'; 67 | private static final String DEFAULT_MODEL = 'gpt-3.5-turbo'; 68 | 69 | @invocableMethod 70 | public static List connect(List inputs) { 71 | // defines the output list 72 | List outputs = new List(); 73 | 74 | for (openAiInputs curInput : inputs) { 75 | String chatHistory = curInput.openaiChatHistory; 76 | String systemPrompt = curInput.openaiSystem; 77 | String prompt = curInput.openaiPrompt; // required 78 | String promptParameters = curInput.promptParameters; 79 | String model = curInput.settingsModel; 80 | Integer tokens = curInput.settingsTokens; 81 | Double temperature = curInput.settingsTemperature; 82 | String endpoint = curInput.settingsEndpoint; 83 | openAiOutputs output = new openAiOutputs(); 84 | String openAiTextOutput = ''; 85 | String openAiMessage = ''; 86 | String finishReason = ''; 87 | String replyID = ''; 88 | Integer tokensUsed = 0; 89 | 90 | // Substitute parameters in prompt 91 | // promptParameters is a JSON string with key/value pairs 92 | if (promptParameters != '' && promptParameters != null) { 93 | Map mapJson; 94 | try { 95 | mapJson = (Map)JSON.deserialize(promptParameters, Map.class); 96 | } 97 | catch (Exception e) { 98 | system.debug('Error parsing Prompt Parameter JSON: ' + e); 99 | } 100 | 101 | for (String key : mapJson.keySet()) { 102 | String value = mapJson.get(key); 103 | prompt = prompt.replace('{' + key + '}', value); 104 | } 105 | } 106 | 107 | if (model == null || model == '') { 108 | model = DEFAULT_MODEL; 109 | } 110 | 111 | if (tokens < 0 || tokens > 4096 || tokens == null) { 112 | tokens = DEFAULT_TOKENS; 113 | } 114 | 115 | if (temperature < 0 || temperature > 1 || temperature == null) { 116 | temperature = DEFAULT_TEMPERATURE; 117 | } 118 | 119 | // Set up the system and user messages 120 | List> messages = new List>(); 121 | 122 | // Add a system message if available 123 | if (systemPrompt != null && systemPrompt != '') { 124 | messages.add(new Map { 125 | 'role' => 'system', 126 | 'content' => systemPrompt 127 | }); 128 | } else if (chatHistory != null && chatHistory != '') { 129 | // Add the chat history 130 | List> chatHistoryList = (List>)JSON.deserialize(chatHistory, List>.class); 131 | 132 | for (Map message : chatHistoryList) { 133 | messages.add(message); 134 | } 135 | } else { 136 | // Add a default system message 137 | messages.add(new Map { 138 | 'role' => 'system', 139 | 'content' => DEFAULT_SYSTEMPROMPT 140 | }); 141 | } 142 | 143 | // Add the user message 144 | messages.add(new Map { 145 | 'role' => 'user', 146 | 'content' => prompt 147 | }); 148 | 149 | // JSON generation for the new chat-based API format 150 | Map jsonBody = new Map { 151 | 'model' => model, 152 | 'messages' => messages, 153 | 'temperature' => temperature, 154 | 'max_tokens' => tokens 155 | }; 156 | String postString = JSON.serialize(jsonBody); 157 | 158 | HttpRequest req = new HttpRequest(); 159 | // set the endpoint to the endpoint defined in named credentials of Salesforce 160 | req.setEndpoint('callout:' + endpoint); 161 | // set the timeout to 3 minutes 162 | req.setTimeout(120000); 163 | // define POST for method endpoint 164 | req.setMethod('POST'); 165 | // set the body of the request 166 | req.setBody(postString); 167 | system.debug('JSON body: ' + req.getBody()); 168 | // create a new HTTP object 169 | Http http = new Http(); 170 | // send the request 171 | HttpResponse res = http.send(req); 172 | // Check the status code of the response 173 | String responseBody = res.getBody(); 174 | output.openAiTemperature = temperature; 175 | 176 | if (res.getStatusCode() == 200) { 177 | System.debug('Response Body: ' + responseBody); 178 | } else { 179 | output.openAiTextOutput = openAiMessage; 180 | System.debug('Error connecting to OpenAI: ' + res.getStatusCode() + '' + res.getStatus() + '' + res.getBody()); 181 | } 182 | 183 | Map deserializedResponse = (Map)JSON.deserializeUntyped(responseBody); 184 | 185 | output.replyID = (String)deserializedResponse.get('id'); 186 | output.openAiObject = (String)deserializedResponse.get('object'); 187 | output.openAiCreated = Datetime.newinstance((Integer)deserializedResponse.get('created')); 188 | output.openAiModel = (String)deserializedResponse.get('model'); 189 | 190 | Map usage = (Map)deserializedResponse.get('usage'); 191 | Integer prompt_tokens = Integer.valueOf(usage.get('prompt_tokens')); 192 | Integer completion_tokens = Integer.valueOf(usage.get('completion_tokens')); 193 | output.tokensUsed = Integer.valueOf(usage.get('total_tokens')); 194 | 195 | List choices = (List)deserializedResponse.get('choices'); 196 | 197 | for (Object choice : choices) { 198 | Map choiceMap = (Map)choice; 199 | Map message = (Map)choiceMap.get('message'); 200 | String role = (String)message.get('role'); 201 | String content = (String)message.get('content'); 202 | String finish_reason = (String)choiceMap.get('finish_reason'); 203 | Integer index = (Integer)choiceMap.get('index'); 204 | output.openAiTextOutput = content; 205 | output.finishReason = finish_reason; 206 | System.debug('Role: ' + role); 207 | System.debug('Content: ' + content); 208 | System.debug('Finish Reason: ' + finish_reason); 209 | System.debug('Index: ' + index); 210 | } 211 | 212 | outputs.add(output); 213 | } 214 | 215 | // return the output list 216 | return outputs; 217 | } 218 | 219 | public class openAiInputs { 220 | @InvocableVariable(required = false label = 'Prompt Parameters' description = 'Is a JSON body of the prompt parameters') 221 | public String promptParameters; 222 | @InvocableVariable(required = false label = 'Open AI System Prompt' description = 'Is the system prompt for what to do with the prompt text') 223 | public String openaiSystem; 224 | @InvocableVariable(required = false label = 'Open AI Chat History' description = 'Is the chat history') 225 | public String openaiChatHistory; 226 | @InvocableVariable(required = true label = 'Open AI Prompt' description = 'Is the prompt for the AI model') 227 | public String openaiPrompt; 228 | @InvocableVariable(required = false label = 'Open AI Model' description = 'Defines the model used of the AI.') 229 | public String settingsModel; 230 | @InvocableVariable(required = false label = 'Tokens' description = 'Maximum Tokens to be used.') 231 | public Integer settingsTokens; 232 | @InvocableVariable(required = false label = 'Temperature' description = 'Temperature of the model to be used.') 233 | public Double settingsTemperature; 234 | @InvocableVariable(required = false label = 'Open AI Endpoint Name' description = 'Named Credential to which you want to call.') 235 | public String settingsEndpoint; 236 | } 237 | 238 | public class openAiOutputs { 239 | @InvocableVariable(required = false label = 'Temperature' description = 'Temperature of the model to be used.') 240 | public Double openAiTemperature; 241 | @InvocableVariable(required = false label = 'Open AI Created' description = 'Date and time the AI was created') 242 | public Datetime openAiCreated; 243 | @InvocableVariable(required = false label = 'Open AI Model' description = 'ID of the AI response') 244 | public String openAiModel; 245 | @InvocableVariable(required = false label = 'Open AI Object Type response' description = 'Is the prompt for the AI model') 246 | public String openAiObject; 247 | @InvocableVariable(required = false label = 'Open AI Text Output' description = 'Provides the text response from OpenAI model as a String.') 248 | public String openAiTextOutput; 249 | @InvocableVariable(required = false label = 'ReplyId' description = 'ReplyID of the Ai response') 250 | public String replyID; 251 | @InvocableVariable(required = false label = 'Finish Reason' description = 'Reason the AI stopped responding') 252 | public String finishReason; 253 | @InvocableVariable(required = false label = 'Tokens Used' description = 'Amount of tokens used.') 254 | public Integer tokensUsed; 255 | } 256 | 257 | } -------------------------------------------------------------------------------- /force-app/main/default/flows/PromptProcessing.flow-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | OpenAI_Connect 5 | 6 | 176 7 | 350 8 | OpenAIConnect 9 | apex 10 | 11 | Increment_Counter 12 | 13 | Automatic 14 | 15 | openaiPrompt 16 | 17 | EditPrompt 18 | 19 | 20 | 21 | openaiSystem 22 | 23 | System_Prompt 24 | 25 | 26 | 27 | promptParameters 28 | 29 | ParameterValues 30 | 31 | 32 | 33 | settingsEndpoint 34 | 35 | NamedCredential 36 | 37 | 38 | 39 | settingsModel 40 | 41 | ModelChoice 42 | 43 | 44 | 45 | settingsTemperature 46 | 47 | Temperature 48 | 49 | 50 | 51 | settingsTokens 52 | 53 | MaxTokens 54 | 55 | 56 | OpenAIConnect 57 | 58 | aiFinishReason 59 | finishReason 60 | 61 | 62 | aiModelUsed 63 | openAiModel 64 | 65 | 66 | alObjectPath 67 | openAiObject 68 | 69 | 70 | TemperatureUsed 71 | openAiTemperature 72 | 73 | 74 | aiAnswer 75 | openAiTextOutput 76 | 77 | 78 | replyId 79 | replyID 80 | 81 | 82 | tokensUsed 83 | tokensUsed 84 | 85 | 1 86 | 87 | 57.0 88 | 89 | NamedCredential 90 | String 91 | 92 | OpenAIChat 93 | 94 | 95 | Execute an OpenAI API call with record data 96 | 97 | Model 98 | Picklist 99 | 100 | 101 | Model__c 102 | Prompt__c 103 | 104 | Default 105 | 106 | AnswerName 107 | String 108 | "Answer for " & {!PromptRecord.Name} & " " & text(now()) 109 | 110 | 111 | IncrementUsedCount 112 | Number 113 | {!PromptRecord.UsedCount__c} + 1 114 | 2 115 | 116 | Prompt Engineering {!$Flow.CurrentDateTime} 117 | 118 | 119 | BuilderType 120 | 121 | LightningFlowBuilder 122 | 123 | 124 | 125 | CanvasMode 126 | 127 | AUTO_LAYOUT_CANVAS 128 | 129 | 130 | 131 | OriginBuilderType 132 | 133 | LightningFlowBuilder 134 | 135 | 136 | Flow 137 | 138 | Record_Answer 139 | 140 | 176 141 | 566 142 | 143 | AnswerScreen 144 | 145 | 146 | Answer__c 147 | 148 | aiAnswer 149 | 150 | 151 | 152 | Name 153 | 154 | AnswerName 155 | 156 | 157 | 158 | ParameterValues__c 159 | 160 | ParameterValues 161 | 162 | 163 | 164 | PromptUsed__c 165 | 166 | EditPrompt 167 | 168 | 169 | 170 | Prompt__c 171 | 172 | PromptRecord.Id 173 | 174 | 175 | 176 | SystemUsed__c 177 | 178 | System_Prompt 179 | 180 | 181 | 182 | Temperature__c 183 | 184 | Temperature 185 | 186 | 187 | 188 | TokensUsed__c 189 | 190 | tokensUsed 191 | 192 | 193 | 194 | Type__c 195 | 196 | Normal 197 | 198 | 199 | PromptAnswer__c 200 | true 201 | 202 | 203 | PromptRecord 204 | 205 | 176 206 | 134 207 | false 208 | 209 | Prompt_Processing 210 | 211 | and 212 | 213 | Id 214 | EqualTo 215 | 216 | promptRecordId 217 | 218 | 219 | true 220 | Prompt__c 221 | true 222 | 223 | 224 | Increment the Used Count field in the prompt 225 | Increment_Counter 226 | 227 | 176 228 | 458 229 | 230 | Record_Answer 231 | 232 | and 233 | 234 | Id 235 | EqualTo 236 | 237 | promptRecordId 238 | 239 | 240 | 241 | UsedCount__c 242 | 243 | IncrementUsedCount 244 | 245 | 246 | Prompt__c 247 | 248 | 249 | AnswerScreen 250 | 251 | 176 252 | 674 253 | true 254 | true 255 | false 256 | 257 | Answer 258 | <p><strong style="font-size: 18px;">ChatGPT Response</strong></p><p><br></p><p>Tokens Used: {!tokensUsed}</p><p>Model Used: {!aiModelUsed}</p><p><br></p><p><u>Answer Text</u></p><p><span style="background-color: rgb(255, 255, 255); color: rgb(68, 68, 68);">{!aiAnswer}</span></p> 259 | DisplayText 260 | 261 | true 262 | true 263 | 264 | 265 | Prompt_Processing 266 | 267 | 176 268 | 242 269 | true 270 | true 271 | false 272 | 273 | OpenAI_Connect 274 | 275 | 276 | Title 277 | <p><strong style="font-size: 14px;">Name: {!PromptRecord.Name}</strong></p><p><span style="font-size: 14px;">Title: {!PromptRecord.Title__c}</span></p> 278 | DisplayText 279 | 280 | 281 | ModelChoice 282 | Model 283 | String 284 | 285 | PromptRecord.Model__c 286 | 287 | Model 288 | DropdownBox 289 | false 290 | 291 | 292 | Temperature 293 | Number 294 | 295 | PromptRecord.Temperature__c 296 | 297 | Temperature (aka Creativity, default = 0.7, up to 1.3 for max) 298 | InputField 299 | false 300 | 2 301 | 302 | 303 | MaxTokens 304 | Number 305 | 306 | PromptRecord.MaxTokens__c 307 | 308 | Max Tokens (default = auto, max = 2000) 309 | InputField 310 | false 311 | 0 312 | 313 | 314 | ParameterValues 315 | String 316 | 317 | PromptRecord.Parameters__c 318 | 319 | Edit parameter values by changing JSON values 320 | InputField 321 | false 322 | 323 | 324 | EditPrompt 325 | 326 | {!PromptRecord.Prompt__c} 327 | 328 | Check and update your prompt if using multi-line parameters 329 | LargeTextArea 330 | false 331 | 332 | 333 | System_Prompt 334 | 335 | {!PromptRecord.System__c} 336 | 337 | System Prompt used to shape response 338 | LargeTextArea 339 | false 340 | 341 | Execute 342 | true 343 | true 344 | 345 | 346 | 50 347 | 0 348 | 349 | PromptRecord 350 | 351 | 352 | Active 353 | 354 | aiAnswer 355 | String 356 | false 357 | false 358 | false 359 | 360 | 361 | aiFinishReason 362 | String 363 | false 364 | false 365 | false 366 | 367 | 368 | aiInstruction 369 | String 370 | false 371 | false 372 | false 373 | 374 | 375 | aiMaxTokens 376 | Number 377 | false 378 | false 379 | false 380 | 0 381 | 382 | 383 | aiModelUsed 384 | String 385 | false 386 | false 387 | false 388 | 389 | 390 | aiPrompt 391 | String 392 | false 393 | false 394 | false 395 | 396 | 397 | alObjectPath 398 | String 399 | false 400 | false 401 | false 402 | 403 | 404 | openAiDate 405 | DateTime 406 | false 407 | false 408 | false 409 | 410 | 411 | promptRecordId 412 | String 413 | false 414 | true 415 | false 416 | 417 | 418 | replyId 419 | String 420 | false 421 | false 422 | false 423 | 424 | 425 | TemperatureUsed 426 | Number 427 | false 428 | false 429 | false 430 | 2 431 | 432 | 433 | tokensUsed 434 | Number 435 | false 436 | false 437 | false 438 | 0 439 | 440 | 441 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # AI Gateway for Salesforce 2 | 3 | ![GitHub release](https://img.shields.io/github/release/vkeenan/sf-prompts.svg) 4 | 5 | Unleash limitless possibilities with our `AI Gateway for Salesforce`, an intuitive Prompt Engineering Platform tailored especially for Salesforce teams. This platform empowers you to create, test, and launch AI prompts within a highly-secured shared repository. The content and responses of these prompts are encapsulated within Salesforce custom objects. With `AI Gateway for Salesforce`, you can leverage parameterized zero-shot prompt engineering to transform your mundane productivity tasks into high-impact efficiencies. 6 | 7 | For detailed insight on how `AI Gateway for Salesforce` works, please refer to our blog post: [Introducing AI Gateway for Salesforce: Managed Access for OpenAI API](https://salesforcedevops.net/index.php/2023/06/15/introducing-ai-gateway-for-salesforce/). 8 | 9 | To install the project, visit Salesforce AppExchange: . You can also leverage this repository for deploying the project to a Salesforce scratch org for testing and development. 10 | 11 | ## Project Description 12 | 13 | `AI Gateway for Salesforce` is a novel, open-source solution that enables Salesforce users to employ the OpenAI API. This tool has been designed with a custom Salesforce object named `Prompt__c`, thereby allowing developers and users to store prompts that can be sent to OpenAI chat completions API using models like `gpt-4` and `gpt-3.5-turbo`. The responses stored securely in `PromptAnswer__c` object can later be retrieved. 14 | 15 | This project presents a minimalist implementation of OpenAI API, serving as a template for organizations to try OpenAI API. The user interface is ingeniously crafted into a simple Flow that administrators and developers can easily modify to ensure customization as per the needs of your organization. 16 | 17 | ![Robot with a Wrench and a Brush](images/SalesforceDevops.net_An_icon_that_is_robot_with_a_wrench.png) 18 | 19 | ## Table of Contents 20 | 21 | - [AI Gateway for Salesforce](#ai-gateway-for-salesforce) 22 | - [Project Description](#project-description) 23 | - [Table of Contents](#table-of-contents) 24 | - [Project Architecture](#project-architecture) 25 | - [Using OpenAI Services](#using-openai-services) 26 | - [Cost Controls using OpenAI Services](#cost-controls-using-openai-services) 27 | - [Privacy Considerations](#privacy-considerations) 28 | - [Personal Identifiable Information](#personal-identifiable-information) 29 | - [OpenAI Cloud Service Provider Considerations](#openai-cloud-service-provider-considerations) 30 | - [Unmanaged Package Post-Installation Notes](#unmanaged-package-post-installation-notes) 31 | - [Loading JSON Sample File](#loading-json-sample-file) 32 | - [Loading CSV Sample File](#loading-csv-sample-file) 33 | - [Local Project Setup](#local-project-setup) 34 | - [OpenAI Setup](#openai-setup) 35 | - [Get OpenAI API Key and Organization ID](#get-openai-api-key-and-organization-id) 36 | - [Salesforce Scratch Org Setup](#salesforce-scratch-org-setup) 37 | - [Add OpenAI Keys to External Credential](#add-openai-keys-to-external-credential) 38 | - [Set Up Permissions for External Credential Principal](#set-up-permissions-for-external-credential-principal) 39 | - [Add Users to Prompt Engineering Permission Set](#add-users-to-prompt-engineering-permission-set) 40 | - [Using AI Gateway for Salesforce](#using-ai-gateway-for-salesforce) 41 | - [Using JSON Parameters](#using-json-parameters) 42 | - [Executing a Prompt](#executing-a-prompt) 43 | - [License](#license) 44 | - [Project Contact Information](#project-contact-information) 45 | - [Project Acknowledgements](#project-acknowledgements) 46 | - [Project Credits](#project-credits) 47 | - [Project Sponsors](#project-sponsors) 48 | - [Project References](#project-references) 49 | 50 | ## Project Architecture 51 | 52 | `AI Gateway for Salesforce` was inspired by the idea "wouldn't it be great if I could share this prompt?" What evolved was a simple solution that allows you to create prompts and share them with other users, but has all the security and safety of Salesforce. 53 | 54 | The workflow is envisioned as follows: 55 | 56 | ![AI Gateway for Salesforce Architecture](images/AI_Gateway_for_Salesforce_Architecture.png) 57 | 58 | Here's a description of that workflow: 59 | 60 | 1. A Prompt Engineer, someone who is knowledgeable about the subject domain and understands how to craft a prompt, creates a parameterized prompt that will perform a certain function. 61 | 2. An end-user then browses a catalog of available prompts. That user selects a prompt and fills out the parameters. The user then executes the prompt. 62 | 3. A Flow is executed that sends the prompt to OpenAI and waits for a response and prints it on the screen. 63 | 64 | ## Using OpenAI Services 65 | 66 | There are a few things you should know to set your expectations about using an OpenAI account. 67 | 68 | 1. OpenAI has many [usage policy restrictions](https://openai.com/policies/usage-policies). 69 | 2. OpenAI is domiciled in the United States and subject to [US laws and export restrictions](https://openai.com/policies/terms-of-use). 70 | 3. You can use the Free Tier of OpenAI API services, but I don't recommend it. You will be limited to the `gpt-3.5-turbo` model and a limited number of API calls per month. 71 | 4. Once you sign up for a paid account, OpenAI will grant you unlimited usage of the `gpt-3.5-turbo` model, but you will need to wait at least 30 days before you can gain access to the `gpt-4` model. 72 | 73 | I set all the sample prompts in the repo to use the `gpt-3.5-turbo` model. You can change the model to `gpt-4` if you have access. 74 | 75 | ### Cost Controls using OpenAI Services 76 | 77 | Using the OpenAI API with GPT4 can lead to unexpected charges if you are not aware of your token usage. OpenAI has billing control features that can put soft and hard stops to your usage once you breach a certain threshold of monthly cost. Go to the following link to define your hard/soft limits for monthly spend once your API key is created () 78 | 79 | GPT consumption is based on the number of tokens used with each request. A good rule of thumb is that every 4 characters will consume one (1) token. OpenAI has an app that can you use to get an estimate on the number of tokens used () 80 | 81 | ## Privacy Considerations 82 | 83 | The enterprise environment brings along critical privacy concerns while using OpenAI API. 84 | 85 | ### Personal Identifiable Information 86 | 87 | It's pivotal to steer clear of storing any personal identifiable information (PII) in the `Prompt__c` object. The `Prompt__c` object isn't encrypted, leaving any data therein unsecured from everyone having access to the object. Similarly, the `PromptAnswer__c` object also doesn't support encryption and hence can't guarantee the safety of PII. 88 | 89 | ### OpenAI Cloud Service Provider Considerations 90 | 91 | You need to make your own decision whether to used OpenAI API because it requires sending data to the Open AI servers. However, if you believe OpenAI's promises, then it should be safe. Here are some of the promises presented in the latest update to their [API data usage policies](https://openai.com/policies/api-data-usage-policies) from June 14, 2023 and the [OpenAI API data privacy](https://openai.com/api-data-privacy) from July 21, 2023: 92 | 93 | 1. OpenAI will not use customer data submitted via the API to train or improve their models unless the customer explicitly opts in to share their data for this purpose. 94 | 2. Data sent through the API will be retained for a maximum of 30 days for abuse and misuse monitoring purposes, after which it will be deleted (unless required by law). 95 | 3. Unlike the consumer ChatGPT-Plus service, where saved conversations are used to improve the model, the API does not save conversations. 96 | 97 | Here is some more information from OpenAI: "By default, OpenAI does not use customer data submitted via the API to train their models or improve their services. Data submitted for fine-tuning is only used to fine-tune the customer's model. OpenAI retains API data for 30 days for abuse and misuse monitoring purposes, and a limited number of authorized employees and third-party contractors can access this data solely for investigating and verifying suspected abuse. Data submitted through the Files endpoint, such as for fine-tuning a model, is retained until the user deletes the file." 98 | 99 | Bottom line: according to these promises your data will not be saved and used for training. But, I would recommend that you read the OpenAI privacy policy and make your own decision about whether you want to use this project in your own org. As of July 25, 2023 we are still waiting for the OpenAI Business account to be available. I believe OpenAI will be able to provide more assurances about data privacy once the Business account is available. 100 | 101 | ## Unmanaged Package Post-Installation Notes 102 | 103 | Don't forget to assign designated users to the `PromptEngineering` permission set! You also need to upload the sample prompts into the `Prompt__c` custom object. You can do this in one of two ways: 104 | 105 | ### Loading JSON Sample File 106 | 107 | You can also load the sample prompts found in this repository to your org. Copy the `Prompt__c.json` document in the `./data` folder to your local machine. Then use the Salesforce CLI to load the data into your org. Use the following command: 108 | 109 | ```bash 110 | sfdx data import tree -f Prompt__c.json --target-org 111 | ``` 112 | 113 | ### Loading CSV Sample File 114 | 115 | If you don't have access to the SDFX or SF commands you can load the sample prompts using a CSV file. I used [dataimporter.io](https://dataimporter.io) to export the sample prompts. To ensure compatibility you should use also use dataimporter.io to load the prompts into your org. You can find the data file in the `./data` folder. 116 | 117 | ## Local Project Setup 118 | 119 | The code provided is designed to work out-of-the-box with Salesforce scratch orgs. You should also be able to adapt it to work with other Salesforce orgs by deploying the metadata elements manually. 120 | 121 | ### OpenAI Setup 122 | 123 | You need to get an OpenAI API key and then configure Salesforce to use it properly. This involves creating a Named Credential and an External Credential. Once we enter the OpenAI keys into Salesforce, they will be safe and cannot be retrieved or copied to your local machine. 124 | 125 | #### Get OpenAI API Key and Organization ID 126 | 127 | To use `AI Gateway for Salesforce` you will need an OpenAI API key and Organization ID. You can get these by following these steps: 128 | 129 | 1. Go to and sign up for an API account. You can get a free tier account, but you will be restricted in the number of api calls you can make. And you'll be restricted to the `gpt-3.5-turbo` model. 130 | 2. Visit and generate a new key for this project. Keep it in a safe place, and don't put it in any repository. 131 | 3. Visit and copy the Organization ID. You'll need this later. 132 | 133 | ### Salesforce Scratch Org Setup 134 | 135 | These scratch org instructions assume you don't already have a `dev-hub` org set up. If you do, you can skip the first two steps. 136 | 137 | 1. You need to have an Salesforce Developer Edition org or create a new one. This will be your `dev-hub` org. You can create a new one here: . 138 | 1. Enable Dev Hub in your `dev-hub` org. See instructions here: . 139 | 1. Install the Salesforce CLI. See instructions here: . 140 | 1. Install VS Code and Salesforce CLI Plugin for VS Code. See instructions here: . 141 | 1. Extract the repository to your local machine. See instructions here: . 142 | 1. Use VS Code to open the `ai-gateway` folder. 143 | 1. Rename the environment file `scripts/.env.example` to `scripts/.env`. 144 | 1. Open the file `scripts/.env` with VS Code and enter your dev-hub username. Save the file. 145 | 1. Open a terminal window in VS Code (Ctrl+Shift+`). 146 | 1. Enter the command `make build` to create the scratch org and load the sample prompts. 147 | 148 | #### Add OpenAI Keys to External Credential 149 | 150 | 1. From the VS Code terminal, enter the command `make open` to go to the scratch org setup page. 151 | 2. Enter "named" in the Quick Find box and select Named Credentials. 152 | 3. Click on the External Credentials tab. 153 | 4. Click on the OpenAI item to edit its details. 154 | 5. In the `Principals` section, click on the down arrow in the `Actions` column next to `PromptEngineering` and select Edit. 155 | 6. Click Add on Authentication Parameters. 156 | 7. Under Parameter 1, enter "apiKey" for the Name. 157 | 8. Enter the OpenAI API key you generated in the previous section for the Value. 158 | 9. Click Add on Authentication Parameters to add your organization id as the second parameter. 159 | 10. Under Parameter 2, enter "Org" for the Name and enter your Organization ID for the Value. 160 | 11. Click Save for to update the external credentials Principals settings. 161 | 162 | #### Set Up Permissions for External Credential Principal 163 | 164 | The External Credential Principal is used to authenticate the external credential. You need to modify the `Prompt Engineering` permission set to gain permission to use it. 165 | 166 | 1. In the Quick Find box, enter "permission" and select Permission Sets. 167 | 2. Click on the `Prompt Engineering` permission set. 168 | 3. Click on `External Credential Principal Access` on the right side of the screen. 169 | 4. Click Edit. 170 | 5. Select the `OpenAI - PromptEngineering` item and click Add. 171 | 6. Click Save. 172 | 173 | #### Add Users to Prompt Engineering Permission Set 174 | 175 | 1. Remain in the `Prompt Engineering` Permission Sets screen from the previous section. 176 | 2. Click on `Manage Assignments`. 177 | 3. Click `Add Assignments`. 178 | 4. Click on the users you want to add to the permission set. 179 | 5. Click `Next`. 180 | 181 | Congratulations! You have completed the setup required to use the OpenAI API from within your Salesforce org, and your API key is safely stored in Salesforce. You are ready to start using the sample prompts. 182 | 183 | ## Using AI Gateway for Salesforce 184 | 185 | Once you have everything set up you can begin sending prompts to OpenAI. Navigate to the Prompt Engineering app from the main application menu. Click the Prompts tab and click New to create a new prompt. Enter a name and title for the prompt and then select the model you want to use. 186 | 187 | You can adjust the temperature setting to control the randomness of the response. The higher the temperature, the more random the response. The lower the temperature, the more predictable the response. The default temperature is 0.7, which is a good starting point. 188 | 189 | You should then enter a System Messaage and User Message. Here is the difference between the two: 190 | 191 | - **System Message**: It should be a message of any length that describes the context of the prompt. You might enter "I am a customer service agent. I am helping a customer with a problem." The system message is automatically sent as the first part of the prompt and generally does not vary between prompts. 192 | - **User Message**: If you have a system message, the user message can be a short message with a request, or it could be a longer message that includes source material. For example, you could enter "Please summarize the sentiment and intent of the customer in this chat session." Then, when you execute this prompt you would paste in the chat session. The system message would be the same, but the user message would change depending on the chat session. 193 | 194 | ### Using JSON Parameters 195 | 196 | Some example prompts use JSON parameters and handlebars notation to insert values into a prompt. For example, the `Market Research` sample prompt uses an extensive System Message to definine how to perform the market research and format the prompt. 197 | 198 | But, notice that the Prompt itself is parameterized with the handlebars format: 199 | 200 | ```text 201 | Create a market analysis for {marketName}. 202 | ``` 203 | 204 | And the Paremters field has the following: 205 | 206 | ```json 207 | { "marketName": "Full Market Name" } 208 | ``` 209 | 210 | When you execute this prompt, the system message will be sent to OpenAI with the parameters replaced with the values from the Parameters field. In this case, if you change "Full Market Name" to something like "Salesforce Devops" you will get a market analysis for Salesforce Devops. 211 | 212 | You can parameterize your own prompts by using the handlebars notation in the prompt and then defining the parameters in the Parameters field. The parameters must be valid JSON. 213 | 214 | The Parameters field is optional. If you don't need to parameterize your prompt, you can leave it blank. It is also not intended to be used to insert very large amounts of text. If you need to insert a large amount of text, you should use paste in the information into the Prompt field instead. 215 | 216 | ### Executing a Prompt 217 | 218 | Once you have created a prompt, you can execute it by clicking the `Execute` button. This will send the prompt to OpenAI and you will have to wait for the response. It can take up to two minutes for a response. 219 | 220 | Once the response is received, it is stored in the `PromptAnswer__c` custom object and displayed in the answer screen. Click `Finish` to finish the Flow session. 221 | 222 | ## License 223 | 224 | This project is licensed under the terms of the MIT license. For more details, see the [LICENSE](./LICENSE) file. 225 | 226 | ## Project Contact Information 227 | 228 | For any inquiries about the project, please reach out to `vern@salesforcedevops.net` or . 229 | 230 | ## Project Acknowledgements 231 | 232 | We would like to thank the following for their contributions and support: 233 | 234 | - Salesforce Community: For their continued support and usage of the project. 235 | - OpenAI: For providing the API used in this project. 236 | - [From Andre Tawalbeh: Tap into the Power of ChatGPT via Flow](https://unofficialsf.com/tap-into-the-power-of-chatgpt-via-flow/) 237 | 238 | ## Project Credits 239 | 240 | - Project Lead: [Vernon Keenan](mailto:vern@salesforcedevops.net) 241 | - Contributors 242 | - [Chris Pearson](https://www.linkedin.com/in/cpearson31/) 243 | 244 | ## Project Sponsors 245 | 246 | We are currently looking for sponsors. If you are interested in sponsoring this project, please contact us. 247 | 248 | ## Project References 249 | 250 | - [Salesforce Developer Documentation](https://developer.salesforce.com/docs/) 251 | - [OpenAI API Documentation](https://platform.openai.com/docs/) 252 | -------------------------------------------------------------------------------- /force-app/main/default/pages/PromptCoverPage.page: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 |

Welcome!

19 |

20 | Welcome image 22 |

23 |

We are glad you are here!

24 |

The Prompt Engineering App enables you to interact with OpenAI API, create and store prompts, and receive AI-generated 25 | responses. Check the statistics above to see the current system usage. Happy prompting! 26 |

27 |
28 |

29 | Unleash limitless possibilities with our 30 | AI Gateway for Salesforce, an intuitive Prompt Engineering Platform tailored especially for Salesforce teams. 31 | This platform empowers you to create, test, and launch AI prompts within a highly-secured shared repository. The 32 | content and responses of these prompts are encapsulated within Salesforce custom objects. With 33 | AI Gateway for Salesforce, you can leverage parameterized zero-shot prompt engineering to transform your 34 | mundane productivity tasks into high-impact efficiencies. 35 |

36 | 37 |

38 | For detailed insight on how 39 | AI Gateway for Salesforce works, please refer to our blog post: 40 | Introducing AI Gateway for Salesforce: Managed Access for OpenAI API. 41 |

42 | 43 |

Project Description

44 | 45 |

46 | AI Gateway for Salesforce is a novel, open-source solution that enables Salesforce users to employ the OpenAI 47 | API. This tool has been designed with a custom Salesforce object named 48 | Prompt__c, thereby allowing developers and users to store prompts that can be sent to OpenAI chat completions API using models 49 | like 50 | gpt-4 and 51 | gpt-3.5-turbo. The responses stored securely in 52 | PromptAnswer__c object can later be retrieved. 53 |

54 | 55 |

56 | This project presents a minimalist implementation of OpenAI API, serving as a template for organizations to try OpenAI API. 57 | The user interface is ingeniously crafted into a simple Flow that administrators and developers can easily modify 58 | to ensure customization as per the needs of your organization. 59 |

60 |

Project Architecture

61 |

62 | AI Gateway for Salesforce was inspired by the idea "wouldn't it be great if I could share this prompt?" What evolved was 63 | a simple solution that allows you to create prompts and share them with other users, but has all the security and 64 | safety of Salesforce. 65 |

66 |

The workflow is envisioned as follows:

67 |

68 | AI Gateway for Salesforce Architecture 70 |

71 |

Here's a description of that workflow:

72 |
    73 |
  1. A Prompt Engineer, someone who is knowledgable about the subject domain and understands how to craft a prompt, creates 74 | a parameterized prompt that will perform a certain function.
    2. 75 |
  2. An end-user then browses a catalog of available prompts. That user selects a prompt and fills out the parameters. 76 | The user then executes the prompt.
    3. 77 |
  3. A Flow is executed that sends the prompt to OpenAI and waits for a response and prints it on the screen.
    4. 78 |
79 | 80 |

Using OpenAI Services

81 |

There are a few things you should know to set your expectations about using an OpenAI account.

82 |
    83 |
  1. OpenAI has many 84 | usage policy restrictions.
    2. 85 |
  2. OpenAI is domiciled in the United States and subject to 86 | US laws and export restrictions.
    3. 87 |
  3. You can use the Free Tier of OpenAI API services, but I don't recommend it. You will be limited to the `gpt-3.5-turbo` 88 | model and a limited number of API calls per month.
    4. 89 |
  4. Once you sign up for a paid account, OpenAI will grant you unlimited usage of the `gpt-3.5-turbo` model, but you 90 | will need to wait at least 30 days before you can gain access to the `gpt-4` model.
    5. 91 |
92 |

I set all the sample prompts in the repo to use the `gpt-3.5-turbo` model. You can change the model to `gpt-4` if you 93 | have access.

94 | 95 |

Cost Controls using OpenAI Services

96 |

97 | Using the OpenAI API with GPT4 can lead to unexpected charges if you are not aware of your token usage. OpenAI has billing 98 | control features that can put soft and hard stops to your usage once you breach a certain threshold of monthly cost. 99 | Go to the following link to define your hard/soft limits for monthly spend once your API key is created 100 | https://platform.openai.com/account/billing/limits 101 |

102 |

103 | GPT consumption is based on the number of tokens used with each request. A good rule of thumb is that every 4 characters 104 | will consume one (1) token. OpenAI has an app that can you use to get an estimate on the number of tokens used 105 | https://platform.openai.com/tokenizer 106 |

107 |

Privacy Considerations

108 | 109 |

110 | The enterprise environment brings along critical privacy concerns while using OpenAI API. 111 |

112 | 113 |

Personal Identifiable Information

114 | 115 |

116 | It's pivotal to steer clear of storing any personal identifiable information (PII) in the `Prompt__c` object. The `Prompt__c` 117 | object isn't encrypted, leaving any data therein unsecured from everyone having access to the object. Similarly, 118 | the `PromptAnswer__c` object also doesn't support encryption and hence can't guarantee the safety of PII. 119 |

120 | 121 |

OpenAI Cloud Service Provider Considerations

122 | 123 |

124 | You need to make your own decision whether to used OpenAI API because it requires sending data to the Open AI servers. However, 125 | if you believe OpenAI's promises, then it should be safe. Here are some of the promises presented in the latest update 126 | to their 127 | API data usage policies from June 14, 2023 and the 128 | OpenAI API data privacy from July 21, 2023: 129 |

130 | 131 |
    132 |
  1. OpenAI will not use customer data submitted via the API to train or improve their models unless the customer explicitly 133 | opts in to share their data for this purpose.
    2. 134 |
  2. Data sent through the API will be retained for a maximum of 30 days for abuse and misuse monitoring purposes, after 135 | which it will be deleted (unless required by law).
    3. 136 |
  3. Unlike the consumer ChatGPT-Plus service, where saved conversations are used to improve the model, the API does not 137 | save conversations.
    4. 138 |
139 | 140 |

141 | Here is some more information from OpenAI: "By default, OpenAI does not use customer data submitted via the API to train 142 | their models or improve their services. Data submitted for fine-tuning is only used to fine-tune the customer's model. 143 | OpenAI retains API data for 30 days for abuse and misuse monitoring purposes, and a limited number of authorized 144 | employees and third-party contractors can access this data solely for investigating and verifying suspected abuse. 145 | Data submitted through the Files endpoint, such as for fine-tuning a model, is retained until the user deletes the 146 | file." 147 |

148 | 149 |

150 | Bottom line: according to these promises your data will not be saved and used for training. But, I would recommend that you 151 | read the OpenAI privacy policy and make your own decision about whether you want to use this project in your own 152 | org. As of July 25, 2023 we are still waiting for the OpenAI Business account to be available. I believe OpenAI will 153 | be able to provide more assurances about data privacy once the Business account is available. 154 |

155 |

Getting Started

156 |

These are the post-intallation instructions for both unmanaged package and scratch org installation.

157 |

Get OpenAPI Key and Org ID

158 |
    159 |
  1. 160 | Go to the OpenAI website and sign up for an API account. You can get a free tier account, but you will be restricted 161 | in the number of API calls you can make. And you'll be restricted to the gpt-3.5-turbo model.
    2. 162 |
  2. 163 | Visit the API keys page and generate a new key for this project. Keep it in a safe place, and don't put it 164 | in any repository.
    3. 165 |
  3. 166 | Visit the Organization settings page and copy the Organization ID. You'll need this later.
    4. 167 |
168 | 169 |

Once you have your API key and Organization ID, you can install the `ai-gateway` package and start using it.

170 | 171 |

Add OpenAI Keys to External Credential

172 |
    173 |
  1. From the VS Code terminal, enter the command `make open` to go to the scratch org setup page.
    2. 174 |
  2. Enter "named" in the Quick Find box and select Named Credentials.
    3. 175 |
  3. Click on the External Credentials tab.
    4. 176 |
  4. Click on the OpenAI item to edit its details.
    5. 177 |
  5. In the `Principals` section, click on the down arrow in the `Actions` column next to `PromptEngineering` and select 178 | Edit. 179 |
    6. 180 |
  6. Click Add on Authentication Parameters.
    7. 181 |
  7. Under Parameter 1, enter "apiKey" for the Name.
    8. 182 |
  8. Enter the OpenAI API key you generated in the previous section for the Value.
    9. 183 |
  9. Click Add on Authentication Parameters to add your organization id as the second parameter.
    10. 184 |
  10. Under Parameter 2, enter "Org" for the Name and enter your Organization ID for the Value.
    11. 185 |
  11. Click Save for to update the external credentials Principals settings.
    12. 186 |
187 | 188 |

Set Up Permissions for External Credential Principal

189 |

The External Credential Principal is used to authenticate the external credential. You need to modify the `Prompt Engineering` 190 | permission set to gain permission to use it.

191 |
    192 |
  1. In the Quick Find box, enter "permission" and select Permission Sets.
    2. 193 |
  2. Click on the `Prompt Engineering` permission set.
    3. 194 |
  3. Click on `External Credential Principal Access` on the right side of the screen.
    4. 195 |
  4. Click Edit.
    5. 196 |
  5. Select the `OpenAI - PromptEngineering` item and click Add.
    6. 197 |
  6. Click Save.
    7. 198 |
199 |

Add Users to Prompt Engineering Permission Set

200 |
    201 |
  1. Remain in the `Prompt Engineering` Permission Sets screen from the previous section.
    2. 202 |
  2. Click on `Manage Assignments`.
    3. 203 |
  3. Click `Add Assignments`.
    4. 204 |
  4. Click on the users you want to add to the permission set.
    5. 205 |
  5. Click `Next`.
    6. 206 |
207 |

Unmanaged Package Post-Installation Notes

208 |

Don't forget to assign designated users to the `PromptEngineering` permission set! You also need to upload the sample 209 | prompts into the `Prompt__c` custom object. You can do this in one of two ways: 210 |

211 | 212 |

Loading JSON Sample File

213 |

You can also load the sample prompts found in this repository to your org. Copy the `Prompt__c.json` document in the 214 | `./data` folder to your local machine. Then use the Salesforce CLI to load the data into your org. Use the following 215 | command: 216 |

217 | 
218 |     sfdx data import tree -f Prompt__c.json --target-org <your-org-alias> 
219 |
220 | 221 |

Loading CSV Sample File

222 |

If you don't have access to the SDFX or SF commands you can load the sample prompts using a CSV file. I used 223 | dataimporter.io to export the sample prompts. To ensure compatibility you should use also use dataimporter.io 224 | to load the prompts into your org. You can find the data file in the `./data` folder. 225 |

226 |

Congratulations! You have completed the setup required to use the OpenAI API from within your Salesforce org, and your 227 | API key is safely stored in Salesforce. You are ready to start using the sample prompts.

228 |

Using AI Gateway for Salesforce

229 |

230 | Once you have everything set up you can begin sending prompts to OpenAI. Navigate to the Prompt Engineering app from the 231 | main application menu. Click the Prompts tab and click New to create a new prompt. Enter a name and title for the 232 | prompt and then select the model you want to use. 233 |

234 |

235 | You can adjust the temperature setting to control the randomness of the response. The higher the temperature, the more random 236 | the response. The lower the temperature, the more predictable the response. The default temperature is 0.7, which 237 | is a good starting point. 238 |

239 |

240 | You should then enter a System Message and User Message. Here is the difference between the two: 241 |

242 |
    243 |
  • 244 | System Message: It should be a message of any length that describes the context of the prompt. You might 245 | enter "I am a customer service agent. I am helping a customer with a problem." The system message is automatically 246 | sent as the first part of the prompt and generally does not vary between prompts
    • 247 |
  • 248 | User Message: If you have a system message, the user message can be a short message with a request, or 249 | it could be a longer message that includes source material. For example, you could enter "Please summarize the 250 | sentiment and intent of the customer in this chat session." Then, when you execute this prompt you would paste 251 | in the chat session. The system message would be the same, but the user message would change depending on the chat 252 | session. 253 |
    • 254 |
255 | 256 |

Using JSON Parameters

257 |

258 | Some example prompts use JSON parameters and handlebars notation to insert values into a prompt. For example, the `Market 259 | Research` sample prompt uses an extensive System Message to define how to perform the market research and format 260 | the prompt. 261 |

262 |

263 | But, notice that the Prompt itself is parameterized with the handlebars format: 264 |
265 | Create a market analysis for {marketName}. 266 |

267 |

268 | And the Parameters field has the following: 269 |
270 | { "marketName": "Full Market Name" } 271 |

272 |

273 | When you execute this prompt, the system message will be sent to OpenAI with the parameters replaced with the values from 274 | the Parameters field. In this case, if you change "Full Market Name" to something like "Salesforce Devops" you will 275 | get a market analysis for Salesforce Devops. 276 |

277 |

278 | You can parameterize your own prompts by using the handlebars notation in the prompt and then defining the parameters in 279 | the Parameters field. The parameters must be valid JSON. 280 |

281 |

282 | The Parameters field is optional. If you don't need to parameterize your prompt, you can leave it blank. It is also not intended 283 | to be used to insert very large amounts of text. If you need to insert a large amount of text, you should paste in 284 | the information into the Prompt field instead. 285 |

286 |

Executing a Prompt

287 |

288 | Once you have created a prompt, you can execute it by clicking the `Execute` button. This will send the prompt to OpenAI 289 | and you will have to wait for the response. It can take up to two minutes for a response. 290 |

291 |

292 | Once the response is received, it is stored in the `PromptAnswer__c` custom object and displayed in the answer screen. Click 293 | `Finish` to finish the Flow session. 294 |

295 | 296 |

License

297 |

298 | This project is licensed under the terms of the MIT license. 299 |

300 | 301 |

Project Contact Information

302 |

303 | For any inquiries about the project, please reach out to me at 304 | vern@salesforcedevops.net or 305 | https://linkedin.com/in/vernonkeenan. 306 |

307 |

Project Acknowledgements

308 |

309 | We would like to thank the following for their contributions and support: 310 |

311 | 318 |

Project Credits

319 |

320 | Project Lead: 321 | Vernon Keenan 322 |
323 |
Project Contributors: 324 |

331 |
332 |

333 |

Project References

334 | 342 | 343 | 344 | 345 | -------------------------------------------------------------------------------- /force-app/main/default/profiles/Admin.profile-meta.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | PromptEngineering 5 | false 6 | true 7 | 8 | 9 | OpenAIConnect 10 | true 11 | 12 | 13 | OpenAIConnectTest 14 | true 15 | 16 | 17 | OpenAIHttpCalloutMock 18 | true 19 | 20 | 21 | PromptStatsController 22 | true 23 | 24 | 25 | PromptStatsControllerTest 26 | true 27 | 28 | false 29 | 30 | false 31 | PromptAnswer__c.Answer__c 32 | false 33 | 34 | 35 | false 36 | PromptAnswer__c.ParameterValues__c 37 | false 38 | 39 | 40 | true 41 | PromptAnswer__c.PromptUsed__c 42 | true 43 | 44 | 45 | false 46 | PromptAnswer__c.Prompt__c 47 | false 48 | 49 | 50 | true 51 | PromptAnswer__c.SystemUsed__c 52 | true 53 | 54 | 55 | true 56 | PromptAnswer__c.Temperature__c 57 | true 58 | 59 | 60 | true 61 | PromptAnswer__c.TokensUsed__c 62 | true 63 | 64 | 65 | false 66 | PromptAnswer__c.Type__c 67 | false 68 | 69 | 70 | true 71 | Prompt__c.MaxTokens__c 72 | true 73 | 74 | 75 | false 76 | Prompt__c.Model__c 77 | false 78 | 79 | 80 | false 81 | Prompt__c.Order__c 82 | false 83 | 84 | 85 | false 86 | Prompt__c.Parameters__c 87 | false 88 | 89 | 90 | true 91 | Prompt__c.PromptCategory__c 92 | true 93 | 94 | 95 | false 96 | Prompt__c.Prompt__c 97 | false 98 | 99 | 100 | false 101 | Prompt__c.System__c 102 | false 103 | 104 | 105 | false 106 | Prompt__c.Tags__c 107 | false 108 | 109 | 110 | true 111 | Prompt__c.Temperature__c 112 | true 113 | 114 | 115 | false 116 | Prompt__c.Title__c 117 | false 118 | 119 | 120 | false 121 | Prompt__c.UsedCount__c 122 | false 123 | 124 | 125 | PromptAnswer__c-Prompt Response Layout 126 | 127 | 128 | Prompt__c-Prompt Layout 129 | 130 | 131 | true 132 | true 133 | true 134 | true 135 | true 136 | PromptAnswer__c 137 | true 138 | 139 | 140 | true 141 | true 142 | true 143 | true 144 | true 145 | Prompt__c 146 | true 147 | 148 | 149 | PromptCoverPage 150 | true 151 | 152 | 153 | Prompt_Engineering 154 | DefaultOn 155 | 156 | 157 | Prompt__c 158 | Hidden 159 | 160 | Salesforce 161 | 162 | true 163 | AIViewInsightObjects 164 | 165 | 166 | true 167 | ActivateContract 168 | 169 | 170 | true 171 | ActivateOrder 172 | 173 | 174 | true 175 | ActivitiesAccess 176 | 177 | 178 | true 179 | AddDirectMessageMembers 180 | 181 | 182 | true 183 | AllowUniversalSearch 184 | 185 | 186 | true 187 | AllowViewKnowledge 188 | 189 | 190 | true 191 | ApexRestServices 192 | 193 | 194 | true 195 | ApiEnabled 196 | 197 | 198 | true 199 | AssignPermissionSets 200 | 201 | 202 | true 203 | AssignTopics 204 | 205 | 206 | true 207 | AuthorApex 208 | 209 | 210 | true 211 | BulkMacrosAllowed 212 | 213 | 214 | true 215 | CanInsertFeedSystemFields 216 | 217 | 218 | true 219 | CanUseNewDashboardBuilder 220 | 221 | 222 | true 223 | CanVerifyComment 224 | 225 | 226 | true 227 | ChangeDashboardColors 228 | 229 | 230 | true 231 | ChatterEditOwnPost 232 | 233 | 234 | true 235 | ChatterEditOwnRecordPost 236 | 237 | 238 | true 239 | ChatterFileLink 240 | 241 | 242 | true 243 | ChatterInternalUser 244 | 245 | 246 | true 247 | ChatterInviteExternalUsers 248 | 249 | 250 | true 251 | ChatterOwnGroups 252 | 253 | 254 | true 255 | ClientSecretRotation 256 | 257 | 258 | true 259 | ConnectOrgToEnvironmentHub 260 | 261 | 262 | true 263 | ConsentApiUpdate 264 | 265 | 266 | true 267 | ContentAdministrator 268 | 269 | 270 | true 271 | ContentWorkspaces 272 | 273 | 274 | true 275 | ConvertLeads 276 | 277 | 278 | true 279 | CreateCustomizeDashboards 280 | 281 | 282 | true 283 | CreateCustomizeFilters 284 | 285 | 286 | true 287 | CreateCustomizeReports 288 | 289 | 290 | true 291 | CreateDashboardFolders 292 | 293 | 294 | true 295 | CreateLtngTempFolder 296 | 297 | 298 | true 299 | CreateReportFolders 300 | 301 | 302 | true 303 | CreateTopics 304 | 305 | 306 | true 307 | CreateWorkBadgeDefinition 308 | 309 | 310 | true 311 | CreateWorkspaces 312 | 313 | 314 | true 315 | CustomizeApplication 316 | 317 | 318 | true 319 | DataExport 320 | 321 | 322 | true 323 | DelegatedTwoFactor 324 | 325 | 326 | true 327 | DeleteActivatedContract 328 | 329 | 330 | true 331 | DeleteTopics 332 | 333 | 334 | true 335 | DistributeFromPersWksp 336 | 337 | 338 | true 339 | EditActivatedOrders 340 | 341 | 342 | true 343 | EditBillingInfo 344 | 345 | 346 | true 347 | EditBrandTemplates 348 | 349 | 350 | true 351 | EditCaseComments 352 | 353 | 354 | true 355 | EditEvent 356 | 357 | 358 | true 359 | EditHtmlTemplates 360 | 361 | 362 | true 363 | EditKnowledge 364 | 365 | 366 | true 367 | EditMyDashboards 368 | 369 | 370 | true 371 | EditMyReports 372 | 373 | 374 | true 375 | EditOppLineItemUnitPrice 376 | 377 | 378 | true 379 | EditPublicDocuments 380 | 381 | 382 | true 383 | EditPublicFilters 384 | 385 | 386 | true 387 | EditPublicTemplates 388 | 389 | 390 | true 391 | EditReadonlyFields 392 | 393 | 394 | true 395 | EditTask 396 | 397 | 398 | true 399 | EditTopics 400 | 401 | 402 | true 403 | EmailMass 404 | 405 | 406 | true 407 | EmailSingle 408 | 409 | 410 | true 411 | EnableCommunityAppLauncher 412 | 413 | 414 | true 415 | EnableNotifications 416 | 417 | 418 | true 419 | ExportReport 420 | 421 | 422 | true 423 | FieldServiceAccess 424 | 425 | 426 | true 427 | GiveRecognitionBadge 428 | 429 | 430 | true 431 | ImportCustomObjects 432 | 433 | 434 | true 435 | ImportLeads 436 | 437 | 438 | true 439 | ImportPersonal 440 | 441 | 442 | true 443 | InstallPackaging 444 | 445 | 446 | true 447 | LightningConsoleAllowedForUser 448 | 449 | 450 | true 451 | LightningExperienceUser 452 | 453 | 454 | true 455 | ListEmailSend 456 | 457 | 458 | true 459 | ManageAnalyticSnapshots 460 | 461 | 462 | true 463 | ManageAuthProviders 464 | 465 | 466 | true 467 | ManageBusinessHourHolidays 468 | 469 | 470 | true 471 | ManageC360AConnections 472 | 473 | 474 | true 475 | ManageCMS 476 | 477 | 478 | true 479 | ManageCallCenters 480 | 481 | 482 | true 483 | ManageCases 484 | 485 | 486 | true 487 | ManageCategories 488 | 489 | 490 | true 491 | ManageCertificates 492 | 493 | 494 | true 495 | ManageContentPermissions 496 | 497 | 498 | true 499 | ManageContentProperties 500 | 501 | 502 | true 503 | ManageContentTypes 504 | 505 | 506 | true 507 | ManageConvMiningReports 508 | 509 | 510 | true 511 | ManageCustomPermissions 512 | 513 | 514 | true 515 | ManageCustomReportTypes 516 | 517 | 518 | true 519 | ManageDashbdsInPubFolders 520 | 521 | 522 | true 523 | ManageDataCategories 524 | 525 | 526 | true 527 | ManageDataIntegrations 528 | 529 | 530 | true 531 | ManageDynamicDashboards 532 | 533 | 534 | true 535 | ManageEmailClientConfig 536 | 537 | 538 | true 539 | ManageEntitlements 540 | 541 | 542 | true 543 | ManageExchangeConfig 544 | 545 | 546 | true 547 | ManageHealthCheck 548 | 549 | 550 | true 551 | ManageHubConnections 552 | 553 | 554 | true 555 | ManageInteraction 556 | 557 | 558 | true 559 | ManageInternalUsers 560 | 561 | 562 | true 563 | ManageIpAddresses 564 | 565 | 566 | true 567 | ManageKnowledge 568 | 569 | 570 | true 571 | ManageKnowledgeImportExport 572 | 573 | 574 | true 575 | ManageLeads 576 | 577 | 578 | true 579 | ManageLoginAccessPolicies 580 | 581 | 582 | true 583 | ManageMobile 584 | 585 | 586 | true 587 | ManageNetworks 588 | 589 | 590 | true 591 | ManageOrchInstsAndWorkItems 592 | 593 | 594 | true 595 | ManagePackageLicenses 596 | 597 | 598 | true 599 | ManagePasswordPolicies 600 | 601 | 602 | true 603 | ManageProfilesPermissionsets 604 | 605 | 606 | true 607 | ManagePropositions 608 | 609 | 610 | true 611 | ManagePvtRptsAndDashbds 612 | 613 | 614 | true 615 | ManageRecommendationStrategies 616 | 617 | 618 | true 619 | ManageReleaseUpdates 620 | 621 | 622 | true 623 | ManageRemoteAccess 624 | 625 | 626 | true 627 | ManageReportsInPubFolders 628 | 629 | 630 | true 631 | ManageRoles 632 | 633 | 634 | true 635 | ManageSearchPromotionRules 636 | 637 | 638 | true 639 | ManageSharing 640 | 641 | 642 | true 643 | ManageSolutions 644 | 645 | 646 | true 647 | ManageSubscriptions 648 | 649 | 650 | true 651 | ManageSynonyms 652 | 653 | 654 | true 655 | ManageUnlistedGroups 656 | 657 | 658 | true 659 | ManageUsers 660 | 661 | 662 | true 663 | MassInlineEdit 664 | 665 | 666 | true 667 | MergeTopics 668 | 669 | 670 | true 671 | ModerateChatter 672 | 673 | 674 | true 675 | ModifyAllData 676 | 677 | 678 | true 679 | ModifyDataClassification 680 | 681 | 682 | true 683 | ModifyMetadata 684 | 685 | 686 | true 687 | NewReportBuilder 688 | 689 | 690 | true 691 | OmnichannelInventorySync 692 | 693 | 694 | true 695 | Packaging2 696 | 697 | 698 | true 699 | Packaging2Delete 700 | 701 | 702 | true 703 | PrivacyDataAccess 704 | 705 | 706 | true 707 | RemoveDirectMessageMembers 708 | 709 | 710 | true 711 | ResetPasswords 712 | 713 | 714 | true 715 | RunReports 716 | 717 | 718 | true 719 | ScheduleReports 720 | 721 | 722 | true 723 | SelectFilesFromSalesforce 724 | 725 | 726 | true 727 | SendCustomNotifications 728 | 729 | 730 | true 731 | SendExternalEmailAvailable 732 | 733 | 734 | true 735 | SendSitRequests 736 | 737 | 738 | true 739 | ShareInternalArticles 740 | 741 | 742 | true 743 | ShowCompanyNameAsUserBadge 744 | 745 | 746 | true 747 | SolutionImport 748 | 749 | 750 | true 751 | SubmitMacrosAllowed 752 | 753 | 754 | true 755 | SubscribeDashboardRolesGrps 756 | 757 | 758 | true 759 | SubscribeDashboardToOtherUsers 760 | 761 | 762 | true 763 | SubscribeReportRolesGrps 764 | 765 | 766 | true 767 | SubscribeReportToOtherUsers 768 | 769 | 770 | true 771 | SubscribeReportsRunAsUser 772 | 773 | 774 | true 775 | SubscribeToLightningDashboards 776 | 777 | 778 | true 779 | SubscribeToLightningReports 780 | 781 | 782 | true 783 | TransactionalEmailSend 784 | 785 | 786 | true 787 | TransferAnyCase 788 | 789 | 790 | true 791 | TransferAnyEntity 792 | 793 | 794 | true 795 | TransferAnyLead 796 | 797 | 798 | true 799 | UseOmnichannelInventoryAPIs 800 | 801 | 802 | true 803 | UseTeamReassignWizards 804 | 805 | 806 | true 807 | UseWebLink 808 | 809 | 810 | true 811 | ViewAllData 812 | 813 | 814 | true 815 | ViewAllProfiles 816 | 817 | 818 | true 819 | ViewAllUsers 820 | 821 | 822 | true 823 | ViewDataAssessment 824 | 825 | 826 | true 827 | ViewDataCategories 828 | 829 | 830 | true 831 | ViewDataLeakageEvents 832 | 833 | 834 | true 835 | ViewDeveloperName 836 | 837 | 838 | true 839 | ViewEventLogFiles 840 | 841 | 842 | true 843 | ViewFlowUsageAndFlowEventData 844 | 845 | 846 | true 847 | ViewHealthCheck 848 | 849 | 850 | true 851 | ViewHelpLink 852 | 853 | 854 | true 855 | ViewMLModels 856 | 857 | 858 | true 859 | ViewMyTeamsDashboards 860 | 861 | 862 | true 863 | ViewPlatformEvents 864 | 865 | 866 | true 867 | ViewPublicDashboards 868 | 869 | 870 | true 871 | ViewPublicReports 872 | 873 | 874 | true 875 | ViewRoles 876 | 877 | 878 | true 879 | ViewSetup 880 | 881 | 882 | true 883 | ViewUserPII 884 | 885 | 886 | true 887 | WorkCalibrationUser 888 | 889 | 890 | -------------------------------------------------------------------------------- /data/Prompt__c.csv: -------------------------------------------------------------------------------- 1 | Max Tokens,Model,Order,Parameters,Prompt,Prompt Category,Prompt Name,System,Tags,Temperature,Title 2 | "","gpt-3.5-turbo","0.0","{""Inquiry"": ""the Inquiry here""}","{Iniquiry}","Devopment","Next.js and Typescript","Context: 3 | You are an AI pair programmer assisting in the creation of a Next.js application in Typescript, utilizing app router as documented on the official Next.js site. The user you are assisting is highly proficient in Typescript, Database Management Systems (DBMS), and REST API technology. The application interfaces with a microservices cluster built in Go through REST APIs for data retrieval. Additionally, the application has a GraphQL interface to an existing WordPress Content Management System (CMS) backend, used to deploy a blog on top of a knowledge management system. 4 | 5 | Problem or Question: 6 | The user is seeking guidance, suggestions, and code snippets on structuring the Next.js application, setting up the app router, and correctly interfacing with the Go-based microservices cluster and the Wordpress CMS backend. The user may have questions about best practices, potential pitfalls, and optimization strategies. 7 | 8 | Relevant Information: 9 | The user is experienced in Typescript and DBMS, which implies a strong understanding of type safety, object-relational mapping, and data modeling. With their expertise in REST API technology, they are familiar with HTTP methods, status codes, and designing and consuming APIs. The application will be communicating with microservices and a CMS backend, indicating a potentially complex data flow and the need for efficient state management. 10 | 11 | Desired Outcome: 12 | The successful outcome of this conversation would be the user gaining a clear understanding of how to structure the Next.js application, use the app router effectively, and establish successful communication with the Go-based microservices and the WordPress CMS. This could be in the form of a detailed walkthrough, code snippets, or addressing specific queries the user may have. 13 | 14 | Tone and Style: 15 | Given the user's experience and the technical nature of the subject, the tone of the conversation should be professional and technical. The style should be detailed, precise, and informative. The responses should include technical jargon relevant to the subject, as the user would be familiar with it. However, the explanations should still be clear and easy to follow. 16 | 17 | Now, rerunning the ACP model in this context, let's address the user's needs effectively and accurately. Any specific questions or areas where you'd like to start, or should we begin with an overview of structuring a Next.js application with app router in Typescript?","","0.7","AI Pair Programmer for Next.js and Typescript Projects - ACP" 18 | "","gpt-3.5-turbo","0.0","{""theCode""}","Refactor this code 19 | ```go 20 | {theCode} 21 | ```","Devopment","Go Refactor","You are acting as a pair programmer for a user working on a project in Go. Your role includes offering advice and best practices for refactoring Go code, identifying common issues and pitfalls, and suggesting performance optimization techniques. 22 | 23 | Your guidance should be based on the principles and features of Go up to version 1.20, including error handling with 'try' and 'catch', the use of generics, and the 'go:embed' directive. However, your focus in this interaction is on refactoring, problem-solving, and performance optimization. 24 | 25 | Go is a statically typed, compiled language that emphasizes simplicity and readability, making it a powerful tool for refactoring. Refactoring in Go is often about simplifying code, improving readability, and enhancing efficiency. 26 | 27 | 1. **Simplifying Code:** Go encourages writing simple, clear code. Unnecessary complexity should be refactored into simpler constructs. For example, if a function has too many arguments, it might be a sign that it's doing too much. Consider breaking it down into smaller, more manageable functions. 28 | 29 | ```go 30 | // Before 31 | func createReport(database *sql.DB, userID, reportID int, startDate, endDate time.Time) (*Report, error) { 32 | //... 33 | } 34 | 35 | // After 36 | type ReportParams struct { 37 | Database *sql.DB 38 | UserID int 39 | ReportID int 40 | StartDate time.Time 41 | EndDate time.Time 42 | } 43 | func createReport(params ReportParams) (*Report, error) { 44 | //... 45 | } 46 | ``` 47 | 48 | 2. **Improving Readability:** Readability is crucial in Go. It's important to follow Go's naming conventions and formatting rules. Use `gofmt` or `goimports` to format your code, and refactor code to make it more readable. For example, instead of nesting `if` conditions, consider using early returns. 49 | 50 | ```go 51 | // Before 52 | func foo(bar int) error { 53 | if bar != 0 { 54 | //... do something 55 | } else { 56 | return errors.New(""invalid bar"") 57 | } 58 | return nil 59 | } 60 | 61 | // After 62 | func foo(bar int) error { 63 | if bar == 0 { 64 | return errors.New(""invalid bar"") 65 | } 66 | //... do something 67 | return nil 68 | } 69 | ``` 70 | 71 | 3. **Enhancing Efficiency:** Go is designed for efficiency, but it's still possible to write inefficient Go code. Look for opportunities to refactor for performance. For example, if you're using the `range` keyword to iterate over a large slice and only need the index, not the value, use the `_` blank identifier to avoid copying the value on each iteration. 72 | 73 | ```go 74 | // Before 75 | for i, v := range largeSlice { 76 | // only using i 77 | } 78 | 79 | // After 80 | for i := range largeSlice { 81 | // no value copy on each iteration 82 | } 83 | ``` 84 | 85 | Common issues in Go include error handling, nil pointer dereference, and concurrency-related issues. Go's explicit error handling requires careful attention to ensure errors are properly handled and not ignored. Nil pointer dereference can occur when an object is not fully initialized before a method is called on it. Concurrency-related issues can happen when multiple goroutines access shared state without synchronization. 86 | 87 | Performance optimization in Go often involves understanding and effectively using Go's memory management, goroutines for concurrent processing, and built-in profiling tools. Remember, premature optimization is the root of all evil. Always measure and identify bottlenecks before optimizing. 88 | 89 | Incorporate this understanding of Go refactoring, common issues, and optimization in your advice to the user.","Go 90 | Refactor","0.7","Go Refactor - ACP" 91 | "","gpt-3.5-turbo","0.0","{""theCode""}","Refactor this code 92 | ```apex 93 | {theCode} 94 | ```","Devopment","Apex Refactor","""You are acting as a pair programmer for a user working on a Salesforce project, specifically focusing on Apex programming. Your role includes offering advice and best practices for refactoring Apex code, identifying common issues and pitfalls, and suggesting performance optimization techniques. 95 | 96 | Apex is a strongly typed, object-oriented programming language that allows developers to execute flow and transaction control statements on the Salesforce platform server in conjunction with calls to the API. Refactoring, problem-solving, and performance optimization are key parts of working effectively with Apex. 97 | 98 | 1. **Simplifying Code:** Apex, being an object-oriented language, encourages the encapsulation of data and operations on that data within objects. If a class or method is becoming too large or complex, consider breaking it down. One common practice in Apex is to use 'Service' classes, which contain business logic, and 'Selector' classes, which contain query logic. 99 | 100 | ```apex 101 | // Before 102 | public class ReportGenerator { 103 | public Report generateReport(Id userId, Date startDate, Date endDate) { 104 | // query and business logic mixed together 105 | } 106 | } 107 | 108 | // After 109 | public class ReportService { 110 | private ReportSelector selector; 111 | public ReportService(ReportSelector selector) { 112 | this.selector = selector; 113 | } 114 | public Report generateReport(Id userId, Date startDate, Date endDate) { 115 | // uses the injected selector to query data 116 | // contains only business logic 117 | } 118 | } 119 | public class ReportSelector { 120 | public List selectReports(Id userId, Date startDate, Date endDate) { 121 | // contains only query logic 122 | } 123 | } 124 | ``` 125 | 126 | 2. **Improving Readability:** It's important to follow Apex's naming conventions and to use clear, descriptive names for variables, methods, and classes. Apex is case-insensitive, but it's a common practice to use CamelCase for class names and camelCase for variable and method names. Comments should be used to explain the 'why', not the 'what'. Refactor code to make it more readable. 127 | 128 | ```apex 129 | // Before 130 | public class rGen { 131 | public Report gR(Id uId, Date sd, Date ed) { 132 | //... 133 | } 134 | } 135 | 136 | // After 137 | public class ReportGenerator { 138 | // Generates a report for the given user and date range. 139 | public Report generateReport(Id userId, Date startDate, Date endDate) { 140 | //... 141 | } 142 | } 143 | ``` 144 | 145 | 3. **Enhancing Efficiency:** Apex code is run on Salesforce's servers, and Salesforce imposes strict governor limits on the use of resources to ensure that no single tenant can monopolize shared resources. It's crucial to understand and optimize around these limits. For example, SOQL queries are a common bottleneck. Bulkify your code to process multiple records at once and reduce the number of SOQL queries. 146 | 147 | ```apex 148 | // Before: SOQL query inside a loop 149 | for (Account account : accounts) { 150 | List contacts = [SELECT Id FROM Contact WHERE AccountId = :account.Id]; 151 | //... 152 | } 153 | 154 | // After: bulkified code 155 | Map accountMap = new Map(accounts); 156 | List contacts = [SELECT Id, AccountId FROM Contact WHERE AccountId IN :accountMap.keySet()]; 157 | for (Contact contact : contacts) { 158 | Account account = accountMap.get(contact.AccountId); 159 | //... 160 | } 161 | ``` 162 | 163 | Common issues in Apex include hitting governor limits, mixing query and business logic, and not handling nulls correctly. Proper error handling is also crucial in Apex, including catching and handling exceptions and using system.assert methods in test classes to catch unexpected behavior.","Apex 164 | Refactor","0.7","Apex Pair Programmer - Refactor Code" 165 | "","gpt-3.5-turbo","0.0","{""content"": ""Press Release Content""}","Analyze this press release and generate bullet point summaries. 166 | ``` 167 | {content} 168 | ```","Marketing","Tech News Writer","You are an AI writing assistant with expertise in generating tech news content. Your role is crafting articles for a corporate and information technology leaders and workers who values high-quality journalism with an objective, fact-based perspective. 169 | 170 | The conversation will focus on analyzing industry trends, company reports, and expert perspectives to synthesize key insights. Ensure your responses demonstrate extensive knowledge of the technology sector, emerging innovations, influential leaders, and current events. 171 | 172 | Provide comprehensive, accurate summaries of the information analyzed. Use a professional but engaging tone with precise, unbiased language reflective of a seasoned news writer. Present logical arguments supported by factual evidence. Seek to inform, explain, and simplify complex concepts for a general audience. 173 | 174 | The ideal outcome is an insightful, well-researched article containing credible sources that educates readers on the tech topic. Recommend any additional sources to consult or experts to interview for a balanced perspective. Make assumptions needed to have a natural conversation from the lens of a diligent tech journalist committed to high-quality, substantive reporting.","","1.0","Generate Tech News Articles" 175 | "","gpt-3.5-turbo","0.0","{""articleText"":""The Article Text""}","Generate 15 Twitter promotional messages for this article 176 | ``` 177 | {articleText} 178 | ```","Marketing","Twitter Promotion","Given the key details, themes, and unique selling points of this particular article, create a set of engaging, creative and catchy promotional Twitter messages that would compel users to click through and read the article. Make sure to use appropriate emojis to add a fun and visually appealing element to your messages. Each message should fit within Twitter's character limit of 280 characters.","Twitter","0.8","Generate Twitter Promotional Messages" 179 | "","gpt-3.5-turbo","0.0","{""articleText"": ""Full Article Text""}","Proofread the following text 180 | ``` 181 | {articleText} 182 | ```","Marketing","Proofreader","You are an AI writing assistant who is a meticulous proofreader adhering to the Associated Press Style Guide. Read the provided text and provide corrections, feedback, and suggestions to enhance its clarity, grammar, punctuation, spelling, and overall style. Your analysis should be presented in a clear and systematic manner. For each paragraph, list the corresponding number and detail any errors found and their corrections. Be thorough and specific in your suggestions to ensure consistency and alignment with the Associated Press Style Guide.","","0.8","Proofread an article using the Associated Press Style Guide" 183 | "","gpt-3.5-turbo","0.0","{""marketName"":""Full Market Name""}","Create a market analysis for {marketName}","Marketing","Market Research","You are an expert B2B market researcher who specializes in enterprise IT products. You are highly skilled at determining the primary factors of analysis which distinguish leaders in that market. Write a narrative that explains why these are the primary competitive factors in the industry. Try to use examples. Create a table which lists the market leaders and how each company rates for each factor of analysis. This type of thinking is called `Market Analysis.`","Market Research","","Perform Market Research" 184 | "","gpt-3.5-turbo","4.0","{""CustomObject__c"":""YourCustomObject"",""NewClassName"":""YourClassNameServices""}","```apex 185 | public with sharing class CourseServices { 186 | public static List getById(String recordId) { 187 | Situs.setsize = 1; 188 | return new List { 189 | new Course(recordId) 190 | }; 191 | } 192 | 193 | public static List getAllCourses(Integer theLimit, Integer theOffset) { 194 | Situs.setsize = System.Database.countQuery('SELECT count() FROM Course__c'); 195 | List sObjects; 196 | List courses = new List(); 197 | try { 198 | sObjects = [ 199 | SELECT Id, UUID__c, CreatedBy.UUID__c, CreatedDate, LastModifiedBy.UUID__c, LastModifiedDate, 200 | Description__c, FullDescription__c, ImageAltText__c, ImageURL__c, Logo__c, 201 | Instructor__r.UUID__c, Price__c, Slug__c, Template__r.UUID__c, Title__c 202 | FROM Course__c 203 | WITH SECURITY_ENFORCED 204 | ORDER BY Id 205 | LIMIT :theLimit 206 | OFFSET :theOffset 207 | ]; 208 | } catch (DmlException e) { 209 | System.debug(LoggingLevel.ERROR, 'An unexpected error has occurred: ' + Utility.errormsg(e)); 210 | } 211 | 212 | for (Course__c theSobject : sObjects) { 213 | courses.add(new Course(theSobject)); 214 | } 215 | 216 | return courses; 217 | } 218 | 219 | } 220 | ```","Devopment","Service Class","Here are custom fields for the `{CustomObject__c}` object. 221 | 222 | ```xml 223 | {CustomObjectXMLFields} 224 | ``` 225 | 226 | Generate a new Apex Service class called `{NewClassName}` according to the following pattern:","Service Functions","0.7","Generate Apex Service Class Static Methods" 227 | "","gpt-3.5-turbo","5.0","{""NewClassName"" : ""NewClassNameResponse""}","Generate a new Apex Response class called `{NewClassName}`","","Response Class","You are a helpful AI Pair programmer expert in Salesforce languages. When given patterns surrounded by triple-backticks you reproduce patterns with new identifiers when asked. Here is the pattern for the Apex Response class named CourseResponse. It is your job to faithfully replace the CourseResponse and different instances of the Course* identifier with the new class name. 228 | ```apex 229 | public with sharing class CourseResponse { 230 | public ResponseMeta meta; 231 | public List data; 232 | public CourseResponse() { 233 | meta = new ResponseMeta(); 234 | data = new List(); 235 | } 236 | 237 | } 238 | ```","Response Class","","Generate an Apex Response Class Wrapper" 239 | "","gpt-3.5-turbo","3.0","{""NewTriggerName"" : ""NewClassNameTrigger""}","Generate a new Apex Trigger called `{NewTriggerName}`","","Trigger","You are a helpful AI Pair programmer expert in Salesforce languages. When given patterns surrounded by triple-backticks you reproduce patterns with new identifiers when asked. This is the Apex Handler Class pattern: 240 | ```apex 241 | trigger CourseTrigger on Course__c (after insert) { 242 | if (Trigger.isAfter && Trigger.isInsert) { 243 | CourseTriggerHandler.afterInsert(trigger.NewMap); 244 | } 245 | } 246 | ```","Trigger","","Generate an Apex Trigger" 247 | "","gpt-3.5-turbo","0.0","{""MarkdownText"":""escapedMarkdown"", ""HeaderLevel"": 2}","Convert this markdown to VisualForce. 248 | ```markdown 249 | {MarkdownText} 250 | ```","Devopment","Markdown to VisualForce","You are a helpful AI Pair programmer expert in Salesforce languages. When given patterns surrounded by triple-backticks you reproduce patterns with new identifiers when asked. When fed Markdown text, you convert the following special characters into their escaped `&` equivalents: `{}\'. 251 | 252 | Here is a pattern we use in VisualForce Pages: 253 | ```VisualForce 254 | 255 |

Welcome!

256 |

257 | 259 |

260 |

We are glad you are here. The Prompt Engineering App enables you to interact with the OpenAI API, create and store 261 | prompts, and receive AI-generated responses. Explore the statistics below to see the current system usage. Happy 262 | coding! 263 |

264 |

Getting Started

265 |

 

266 |

Get OpenAPI Key and Org ID

267 |
    268 |
  1. 269 | Go to the OpenAI website and sign up for an API account. You can get a free tier account, but you will be restricted 270 | in the number of API calls you can make. And you'll be restricted to the gpt-3.5-turbo model.
    2. 271 |
  2. 272 | Visit the API keys page and generate a new key for this project. Keep it in a safe place, and don't put it 273 | in any repository.
    3. 274 |
  3. 275 | Visit the Organization settings page and copy the Organization ID. You'll need this later.
    4. 276 |
277 | 278 |

Once you have your API key and Organization ID, you can install the sf-prompts package and start using it.

279 | 280 |

More Information: 281 | https://salesforcedevops.net/prompt-engineering-app 282 |

283 | 284 | ```","VisualForce 285 | Development","0.7","Convert Markdown text into VisualForce HTML Code" 286 | "","gpt-3.5-turbo","2.0","{""NewClassName.cls"" : ""NewClassName.cls""}","Generate a new Apex Trigger Handler class called `{NewClassName.cls}` that follows the following pattern: 287 | 288 | ```apex 289 | public with sharing class CourseTriggerHandler { 290 | public static void afterInsert (Map newMap) { 291 | doUUID(newMap.keySet()); 292 | } 293 | 294 | public static void afterUpdate (Map newMap, Map oldMap) { 295 | doUUID(newMap.keySet()); 296 | } 297 | 298 | private static void doUUID(Set recordIds) { 299 | Boolean updateFlag = false; 300 | List theSobjects = [ 301 | SELECT 302 | Id, 303 | UUID__c 304 | FROM Course__c 305 | WHERE Id IN : recordIds 306 | WITH SECURITY_ENFORCED 307 | ]; 308 | 309 | for (Course__c theSobject : theSobjects) { 310 | if (theSobject.UUID__c == null) { 311 | updateFlag = true; 312 | theSobject.UUID__c = Utility.newUuid(); 313 | } 314 | } 315 | 316 | if (updateFlag) { 317 | Update theSobjects 318 | ; 319 | } 320 | } 321 | 322 | } 323 | ```","Devopment","Trigger Handler Class","You are a helpful AI pair programmer who specializes in Salesforce languages.","Trigger Handler","0.7","Generate Trigger Handler Class" 324 | "","gpt-3.5-turbo","1.0","{""NewClassName.cls"" : ""NewClassName.cls""}","Generate a new Apex Service Wrapper class called `{NewClassName.cls}` that contains the following custom fields: 325 | 326 | ```xml 327 | {NewClassNameFields} 328 | ``` 329 | 330 | Generate {NewClassName} using the `Course` class definition as a template. 331 | 332 | ```apex 333 | public inherited sharing class Course { 334 | public String ID; 335 | public String CreatedById; 336 | public DateTime CreatedDate; 337 | public String Description; 338 | public String FullDescription; 339 | public String ImageAltText; 340 | public String ImageUrl; 341 | public String InstructorId; 342 | public String LastModifedById; 343 | public DateTime LastModifiedDate; 344 | public String Logo; 345 | public Decimal Price; 346 | public String Slug; 347 | public String TemplateId; 348 | public String Title; 349 | 350 | public Course(Course__c theSobject) { 351 | unMarshalDB(theSobject); 352 | } 353 | 354 | public Course(String recordId) { 355 | if (recordId != null) { 356 | String whereClause = (recordId.length() == 36) ? 'UUID__c = :recordId' : 'Id = :recordId'; 357 | String queryString = 'SELECT Id, CreatedBy.UUID__c, CreatedDate, LastModifiedBy.UUID__c, LastModifiedDate, UUID__c, Description__c, FullDescription__c, ImageAltText__c, ImageURL__c, Logo__c, Instructor__r.UUID__c, Price__c, Slug__c, Template__r.UUID__c, Title__c FROM Course__c WHERE ' + whereClause + ' WITH SECURITY_ENFORCED LIMIT 1'; 358 | List sObjList = new List(); 359 | 360 | try { 361 | sObjList = System.Database.query(queryString); 362 | } catch (DmlException e) { 363 | System.debug('An unexpected error has occurred: ' + Utility.errormsg(e)); 364 | } 365 | 366 | if (sObjList.size() == 0) { 367 | System.debug('Course.constructor: unknown Course__c ID: ' + recordId); 368 | } else { 369 | unMarshalDB(sObjList[0]); 370 | } 371 | } 372 | } 373 | 374 | private void unMarshalDB(Course__c theSobject) { 375 | ID = theSobject.UUID__c; 376 | CreatedById = theSobject.CreatedBy.UUID__c; 377 | CreatedDate = theSobject.CreatedDate; 378 | Description = theSobject.Description__c; 379 | FullDescription = theSobject.FullDescription__c; 380 | ImageURL = theSobject.ImageURL__c; 381 | ImageAltText = theSobject.ImageAltText__c; 382 | InstructorId = theSobject.Instructor__r.UUID__c; 383 | LastModifedById = theSobject.LastModifiedBy.UUID__c; 384 | LastModifiedDate = theSobject.LastModifiedDate; 385 | Logo = theSobject.Logo__c; 386 | Price = theSobject.Price__c; 387 | Slug = theSobject.Slug__c; 388 | TemplateId = theSobject.Template__r.UUID__c; 389 | Title = theSobject.Title__c; 390 | } 391 | 392 | } 393 | ```","Devopment","Service Wrapper Class","You are a helpful AI pair programmer who specializes in Salesforce languages.","Service Class Wrapper","","Generate a Service Wrapper Class" 394 | "","gpt-3.5-turbo","6.0","{""NewClassName.cls"" : ""NewClassName.cls""}","Generate a new Apex Handler class called `{NewClassName.cls}`.","","Handler Class","You are a helpful AI Pair programmer expert in Salesforce languages. When given patterns surrounded by triple-backticks you reproduce patterns with new identifiers when asked. This is the Apex Handler Class pattern: 395 | ```apex 396 | @RestResource(urlMapping = '/courses') 397 | global with sharing class HttpCourse { 398 | @HttpGet 399 | global static void getCourses() { 400 | Context.isRest = true; 401 | Datetime queryStart = Datetime.now(); 402 | CourseResponse response = new CourseResponse(); 403 | Integer theOffset = HttpHelper.getOffset(); 404 | Integer theLimit = HttpHelper.getLimit(); 405 | String courseId = RestContext.request.params.get('courseId'); 406 | String altId = RestContext.request.params.get('id'); 407 | 408 | if (altId != null) { 409 | courseId = altId; 410 | } 411 | 412 | if (courseId != null) { 413 | response.data = CourseServices.getById(courseId); 414 | } else { 415 | response.data = CourseServices.getAllCourses(theLimit, theOffset); 416 | } 417 | 418 | // construct response metadata 419 | response.meta.pagination.plimit = theLimit; 420 | response.meta.pagination.poffset = theOffset; 421 | response.meta.pagination.pagesize = response.data.size(); 422 | response.meta.pagination.setsize = situs.setsize; 423 | Datetime queryEnd = Datetime.now(); 424 | Long theTime = (queryEnd.getTime() - queryStart.getTime()); 425 | response.meta.serverresponsetime = String.valueOf(theTime) + ' miliseconds'; 426 | 427 | RestResponse res = RestContext.response; 428 | try { 429 | res.responseBody = Blob.valueOf(JSON.serialize(response)); 430 | res.addHeader('Content-Type', 'application/json'); 431 | res.statusCode = 200; 432 | } catch (Exception e) { 433 | res.responseBody = Blob.valueOf(Utility.errormsg(e)); 434 | res.statusCode = 500; 435 | } 436 | } 437 | 438 | } 439 | ```","Handler Class","","Generate a Handler Class" 440 | "","gpt-3.5-turbo","0.0","{""PressRelease"":""Content goes here""}","Write a summary of this press release: 441 | ``` 442 | {PressRelease} 443 | ```","Marketing","B2B Writer","You are an AI writing assistant equipped with advanced language understanding and text generation skills, specifically trained in the domain of B2B content marketing. The role of B2B content marketing is to build and strengthen relationships with other businesses by providing valuable content that resonates with their problems, needs, and goals. Key strategies in this area include understanding the business customers' persona, creating engaging content that aligns with their journey, leveraging SEO for better visibility, and constantly measuring and optimizing the performance of the content. 444 | 445 | A user is seeking assistance in generating relevant and engaging B2B content for promoting their software service in the IT sector. They intend to highlight the unique features of their service, showcase its benefits to businesses in the IT sector, and create a compelling call-to-action that drives businesses to try or purchase their service. The content will be used across different channels, including their blog, social media platforms, and email marketing campaigns. 446 | 447 | The user wants to leverage your B2B content marketing skills to come up with strategies, create a content plan, and generate engaging content that aligns with their goals and targets. The outcome they desire is to improve brand awareness, increase their service's visibility in the market, attract potential business customers, and boost their sales. 448 | 449 | The user expects an in-depth and strategic discussion, focusing on the specifics of B2B content marketing for services in the IT sector. You should incorporate your knowledge of B2B content marketing principles, practices, and trends into the conversation. Your responses should be clear, logical, and comprehensive, and they should provide actionable advice. The tone of the conversation should be professional and informative, but also engaging, to make the content planning and creation process easy to understand and implement.","Marketing","1.0","B2B Writing Assistant Prompt - ACP" 450 | "","gpt-3.5-turbo","0.0","{""articleText"": ""the article text""}","Generate 10 image descriptions from this: 451 | ``` 452 | {articleText} 453 | ```","Marketing","Midjourney Prompts","Given the narrative and details in the article provided, imagine you are an observer surrounded by the primary elements of the article. Generate a highly detailed, stylistic written depiction of the article. Incorporate all the relevant elements, characters, surroundings, and feelings involved. When an element of the depiction is a trade name, acronym, or company name, create a descriptive avatar representing that element. Do not use tradenames, acronyms, or company names in the image description. These descriptions will be suitable to be a prompt input to a LLM image generator like Midjourney, Stable Diffusion, or DALL-E. 454 | 455 | Let your description paint a vivid picture that reflects the article's unique mood, atmosphere, and style. Follow these guidelines when generating the image description: 456 | * Prefer interior scenes to outdoor landscapes 457 | * You are generating a static image. Do not make references to what is happening before or after the imagined scene. 458 | * Try to make each description 75 to 150 tokens long. 459 | * When a company or product name is included in the article, analyze that entity and assign an appropriate avatar to that entity. 460 | 461 | This is a top priority: Whatever you do, do not refer to the company, concept name, or acronym, or product name in the description. You must come up with an descriptive and relevant character or avatar description. 462 | 463 | Randomly use these Midjourney-specific styles: 464 | 465 | Style Description 466 | 8k Lighting tends to be more extreme; colors even more saturated and computer-generated looking than ""high definition"" 467 | cinematic Shadows tend to be more extreme (though not darker); objects a bit thicker; more poster-like 468 | high definition Shadows are lightened; more fanciful and saturated colors 469 | ultra photorealistic Similar to ""fine ultra-detailed realistic"" 470 | Hasselblad H6D Sharper focus on subject; shadows are deepened 471 | fine ultra-detailed realistic Can be a bit grainy and ""ropey"" but increases detail generation 472 | color grading Extreme variations in hue; vibrant but not over-saturated colors 473 | depth of field Sharp focus on subject, background and foreground blurred 474 | film lighting Limited lighting sources; backlighting common; deep shadows cast by light sources 475 | rim lighting Slightly stronger lighting effect than ""film lighting,"" but very similar results 476 | intricate Tends toward non-realistic ""crafts"" and ""pattern"" type designs. 477 | realism Artistic realism. Backgrounds tend to be more uniform; subject looks more like a painting; more objects with subject 478 | photography Subject tends to have a little area of objects around it with little else in the background 479 | rendered for IMAX More complex subject with very directional lighting and subdued saturation 480 | tilt-shift Like ""depth of field,"" but from above or with high angle 481 | motion-blur Speed lines. May render as if wind is blowing 482 | 35mm film More vibrant colors, but muted saturation, detailed with additional foreground and/or background elements","Image Generation","1.0","Generate Midjourney Prompts from Article Text - ACP" 483 | "","gpt-3.5-turbo","0.0","{""articleText"":""The Article Text""}","Generate 15 LinkedIn promotional messages for this article 484 | ``` 485 | {articleText} 486 | ```","Marketing","LinkedIn Promotion","Reflecting the key points, professional relevance, and unique insights of the provided article, create a series of engaging and enticing LinkedIn posts aimed to draw in and connect with your professional network. Your posts should motivate LinkedIn users to click through and read the full article. Maintain a tone that's professional yet engaging, and enhance your posts with suitable emojis to add a touch of personality and visual appeal. Be sure to include industry-specific keywords and hashtags for increased visibility. Use about one full paragraph of text for each message.","Twitter","0.8","Generate LinkedIn Promotional Messages" 487 | "","gpt-3.5-turbo","0.0","{""subject"": ""System Prompt Subject""}","Generate a system prompt using ACP for {subject}.","Devopment","Advanced Contextual Preamble","You are an advanced AI language model trained to understand and generate human-like text based on the input you receive. Your core strength lies in your ability to synthesize information, answer questions, and generate creative content. Now, you're being asked to assist in the creation of Advanced Contextual Preamble (ACP) prompts. ACP is a technique where the user leverages the system prompt to provide high-level instructions or context to guide the model's responses effectively. Here are the key principles to remember when creating ACP prompts: 488 | 1. Establish the Context: Begin by providing an overview of the context within which the user's question or problem exists. This might involve explaining the subject, the setting, or the key concepts that will be relevant in the conversation. 489 | 2. Highlight the Problem or Question: Once the context is set, briefly state the question or problem that needs to be addressed. This gives a direction to the conversation and helps you understand the user's intent. 490 | 3. Share Relevant Information: Include any details, facts, or data points that might be relevant to the problem or question. This could include information that is not directly related but might influence the conversation or the solutions that are provided. 491 | 4. Define the Desired Outcome: Explain what a successful outcome of the conversation would look like. This could be an answer to a question, a solution to a problem, or simply a meaningful and engaging conversation on a given topic. 492 | 5. Set the Tone and Style: Finally, indicate the desired tone and style of the conversation. Whether the conversation should be formal or informal, technical or layman, serious or humorous, setting the tone helps in aligning the responses to the user's expectations. 493 | 6. Create a Lengthy Response: ACP prompts should be at least 1,500 tokens in length to create an extensive context. 494 | Remember, the goal of an ACP is to use the system prompt to guide the model's responses by providing as much context and information as possible. The more detailed and specific the context, the more effective the model can be in providing relevant, accurate, and useful responses. When asked to “rerun the ACP model” you are asked to reconsider the above factors in crafting a prompt designed to shape accurate responses in a certain subject. 495 | Now, let's get started. You are helping a user craft the idea prompt for starting a conversation with an LLM on a specific topic, like architecting an IT solution, managing a team, starting a business, coding in a specific programming language, or learning a new skill. You will be asked questions limited to business management, IT service management, entrepreneurship, software engineering, distributed computing system architecture, SaaS system implementation, devops, platform engineering, and enterprise application development. Before giving a response that fits the ACP model, add any relevant knowledge to your context and re-run the ACP model in this context to guide your responses. Always strive to provide the most relevant, accurate, and helpful information based on the context and information provided in the conversation.","Prompting","0.7","Generate Prompts with Advanced Contextual Preamble" 496 | "","gpt-3.5-turbo","0.0","{""goTask"":""Full description""}","TASK: {goTask}","Devopment","Go Pair Programmer","You are an AI language model with deep expertise in the Go programming language, OpenAPI 2.0 and 3.0, and proficiency in the use of the go-swagger package and openapi-generator-cli for multiple languages. You're tasked with assisting the user in a pair programming setting, where you'll be working on projects involving API design, code generation, API testing, and developing Go applications. Your areas of expertise include: 497 | 498 | 1. **Go Programming**: You're proficient with the Go programming language, familiar with its syntax, standard libraries, conventions, and ecosystem. You can assist with best practices in Go coding, unit testing, benchmarking, and debugging. You understand and can explain concepts like goroutines, channels, defer, panic & recover, and interfaces in Go. 499 | 500 | 2. **OpenAPI Specifications**: You have a detailed understanding of OpenAPI specifications, both 2.0 (formerly known as Swagger) and the newer 3.0 version. You're well-versed in the structure of OpenAPI documents including paths, operations, parameters, request bodies, responses, and components. 501 | 502 | 3. **go-swagger package**: You can guide the user through the process of generating, building, and deploying a server or client in Go from an OpenAPI specification using go-swagger. 503 | 504 | 4. **openapi-generator-cli**: You can assist in generating API client libraries, server stubs, API documentation, and configuration files in a variety of languages from OpenAPI specifications using openapi-generator-cli. 505 | 506 | 5. **API Design and Testing**: You can help the user design robust, scalable, and user-friendly APIs, including advice on versioning, error handling, pagination, filtering, sorting, and other common API features. Additionally, you understand various testing methodologies for APIs, such as unit tests, integration tests, and end-to-end tests. 507 | 508 | 6. **Integration and Security Practices**: You can assist the user in integrating their APIs with databases, external APIs, authentication services, and other back-end services. You can also guide the user in implementing standard security practices for APIs, such as authentication, authorization, rate limiting, and securing sensitive data. 509 | 510 | 7. **CI/CD**: You can help the user set up a CI/CD pipeline for their API project, including automated building, testing, and deployment. 511 | 512 | Your primary objective is to be a proficient pair programming partner, aiding the user in designing and implementing robust APIs, understanding OpenAPI specifications, improving their Go programming skills, and enhancing their overall software engineering practices. Your responses should be informative, clear, and user-focused, and you should aim to understand the user's perspective and cater to their specific needs.","","0.1","Go Pair Programmer - ACP" 513 | --------------------------------------------------------------------------------