Skip to content

Team_Playbook.md

Latest commit

 

History

History
48 lines (38 loc) · 2.17 KB

Team_Playbook.md

File metadata and controls

48 lines (38 loc) · 2.17 KB
Raw

Agentforce Data‑Aware Solutions — Team Playbook

This playbook explains how to set up, deploy, operate, and extend the Agentforce data‑aware agent templates.

What you get

Two complementary Salesforce DX projects demonstrating advanced AI agent capabilities:

1. Agentforce Data-Aware Agent (Lead Qualification)

  • Auto‑discovery of org schema (objects/fields/relationships) with FLS/sharing awareness
  • SchemaGraph__c cache for fast lookups
  • Planner + Plugin + Functions so agents can find/use data without manual mapping
  • Guardrails: FLS enforcement, RestrictedField__mdt deny‑list, audit stubs
  • Lead Qualification demo (Flow + Apex invocable)
  • Unit tests and a GitHub Actions PR validation pipeline

2. Personal Shopping Assistant (E-commerce Bot)

  • Dynamic product catalog discovery and inventory management
  • Real-time stock checking and personalized recommendations
  • Customer loyalty integration and shopping cart management
  • Knowledge-grounded responses using Salesforce Knowledge
  • Einstein Bot interface for customer interactions

Prereqs

  • Dev Hub with Scratch Orgs enabled, Salesforce CLI, GitHub repo, Connected App for JWT

Create scratch & push

sf org login web --alias devhub --set-default-dev-hub
sf org create scratch --definition-file config/project-scratch-def.json --alias agent-data-aware --set-default --duration-days 7 --wait 10
sf project deploy start --ignore-conflicts --wait 30
sf org assign permset --name GenAIAgentPermission
sf apex run --file scripts/apex/run_bootstrap.apex
sf org open

Lead Qualification demo

Ask the bot: “Qualify this lead and follow up: 00Qxxxxxxxxxxxx; set status to Working”. Planner → FindFields → ExecuteSOQL → Flow (status) → Apex (task).

CI (PRs)

JWT login → create scratch → deploy → run tests (JUnit) → upload results → delete org. Secrets: SF_CONSUMER_KEY, SF_USERNAME, SF_JWT_KEY.

Operating

  • Nightly schema refresh: OrgSchemaNightlyJob.scheduleNightly()
  • On‑demand rebuild: OrgSchemaBootstrap.buildSchemaGraph()
  • Add sensitive fields to RestrictedField__mdt; re‑run bootstrap

Extending

  • Add new GenAiFunction (Apex/Flow), list in Planner <actions>, add tests, open PR.