Agentforce Specialist Study Guide
The CertifyForce Agentforce Specialist study guide covers the official Salesforce Agentforce Specialist exam outline in 26 units across all 6 exam sections, one unit per exam-guide topic. The exam itself has 60 scored questions, a 105-minute limit and a 72% passing score, and its largest section is AI Agents at 35%. Every unit is verified against official Salesforce documentation, cites its sources, and links to the practice questions that test it. Last updated October 15, 2026.
The Salesforce Agentforce Specialist credential is designed for individuals who are responsible for managing Agentforce within their organizations. The Agentforce Specialist has a deep understanding of both Agentforce capabilities and Salesforce Platform configuration. This exam is designed for individuals looking to validate their expertise in implementing Agentforce to drive business success.
26 lessons · 6 exam sections · about 3 h of reading ·
Exam sections
Unlock the full guide
Every lesson's summary, exam expectations, key facts and self-check questions are free to read. The Agentforce Specialist pack unlocks the self-check answers, core concepts, exam traps and worked scenarios, reading progress, personalized “read next”, and unlimited practice exams.
Prompt Engineering
20% of the exam- Prompt Builder use cases and solution fit~8 minWhen a requirement calls for a Prompt Builder template (generated language grounded in Salesforce data that admins can tune) and when it calls for a formula, roll-up summary, flow or validation rule because the result must be exact, repeatable and auditable.
- Prompt template permissions and access controls~7 minWho can build, activate and run prompt templates (Prompt Template Manager, Prompt Template User or Customize Application), and why the executing user's object access, sharing and field-level security decide which data a grounded prompt actually contains.
- Prompt template types: field generation and flex~8 minWhat each prompt template type is bound to, with the considerations the exam tests for field generation (one object, one writable text field, a reviewed Einstein draft on a Dynamic Forms record page) and flex (up to five inputs you define, any caller), and how to pick the type from a requirement.
- Grounding techniques: merge fields, flows and retrievers~8 minHow to choose between record and related list merge fields, template-triggered prompt flows, Apex and Data 360 retrievers to ground a prompt template, and how to wire a retriever into a template, including activation, run-time search settings, dynamic pre-filters and citations.
- Prompt template lifecycle: build, preview, activate, invoke~8 minHow a prompt template goes from New Prompt Template in Setup through the Prompt Template Workspace, Preview, saved versions and activation to execution from a record page, a flow, Apex or an agent, and why an unactivated template is invisible to every consumer.
- Writing effective prompt template instructions~8 minThe documented best practices for prompt template text: a role with a goal, an Instructions: section in triple quotes, explicit format and length, one consistent plain-language style, an instruction to generate only the content, staying inside the grounded data, and iterating with real records and user feedback.
- Einstein Trust Layer security and privacy~8 minWhat each Einstein Trust Layer feature does on the prompt and response journey (secure data retrieval, data masking, prompt defense, the TLS-secured LLM gateway, zero data retention, toxicity scoring, demasking and the audit trail in Data 360), and the exam's favourite exception: data masking is disabled for agents.
- Model access controls and bring your own LLM~7 minHow model use is governed: the model chosen per prompt template version, org-level provider blocking in Setup, hiding an LLM configuration in AI Models (formerly Einstein Studio), beta-model enablement, limiting who holds Manage Prompt Templates, Salesforce's AI policies, and connecting a contracted provider with bring your own LLM.
Data 360 Fundamentals
20% of the exam- Agentforce Data Library setup and considerations~8 minHow an Agentforce Data Library turns Knowledge articles or uploaded files into a working RAG setup: what Salesforce provisions in Data 360, who builds a library and where, file limits, the locked source type, connecting it to an agent, and when to outgrow it with hand-built retrievers.
- Chunking, search indexes and retrievers in Data 360~7 minHow Data 360 makes unstructured content retrievable: chunking and vectorization, vector, hybrid and enriched search indexes, data-space scoping, and individual and ensemble retrievers with filters, versions, citations and the Retriever Playground.
AI Agents
35% of the exam- Agent Script building blocks and the agent graph~8 minWhat Agent Script is and which blocks define an agent (system, config, variables, start_agent, subagents), how the script compiles into an Agent Graph that the Atlas Reasoning Engine runs, and what belongs in before_reasoning and after_reasoning.
- Hybrid reasoning in Canvas and Script view~7 minHow Canvas view and Script view edit one Agent Script, where hybrid reasoning draws the line between deterministic code and LLM calls, which lifecycle points spend an LLM call, and when a workflow should stay LLM-led.
- Deterministic control with variables, filters and expressions~8 minHow custom, context and system variables hold agent state, why gate variables default closed and must be set by code, how available-when filters hide actions and subagents, and how conditionals, template expressions and transitions enforce a workflow without action loops.
- Standard and custom topics and actions~8 minWhen to use a standard or custom subagent (formerly topic) and a standard or custom agent action, which reference types a custom action can call, how action instructions and input and output settings shape behaviour, how Agentforce Assets copies work, and how versions and activation govern editing a live agent.
- Agent channels: messaging, voice, email and Slack~8 minHow connections in Agentforce Builder put one agent on many channels: Enhanced Chat and messaging apps with Omni-Channel routing and escalation, Experience Cloud sites, Service Email, Agentforce Voice, and the employee surfaces (Agentforce panel, mobile, Slack), plus adaptive response formats.
- Agent user, run context and private actions~7 minWhose permissions an agent uses: the Service agent's agent user, the logged-in user for Employee agents and authenticated site users, flow and Apex run context inside actions, and how verified IDs, session identifiers and filters keep private actions scoped to one customer.
- Choosing an Employee or Service agent~7 minHow to pick the agent type from the audience, channel and running user: Agentforce Employee agents for staff, Agentforce Service agents for customers, the permissions to build each, Flex Credits billing, migrating retired Agentforce (Default) agents, and splitting two audiences into two agents.
- Agent API for headless agent integrations~6 minWhen to reach an agent over the Agent API instead of a channel, SDK or flow action, and how an integration works: external client app with client credentials, sessions with a session key, sequence IDs, synchronous versus streaming replies, bypassUser, and the documented limits.
Testing, Deployment, and Maintenance
10% of the exam- Testing agents in Testing Center~8 minHow to batch test an agent in Testing Center: why it runs in a sandbox, who can use it, turn-level versus conversation-level suites, generating or uploading test cases, the CSV template's API-name rules, and the limits that shape a large test plan.
- Testing Center evaluations and test results~7 minWhat each Testing Center evaluation compares, how default, quality and custom scorers are judged and scored, why non-deterministic agents still need evaluation at scale, and how to read a run and act on failed test cases.
- Deploying agents from sandbox to production~8 minMoving a tested agent to production with a change set or with Salesforce CLI and source control: which metadata types carry the agent, why every dependency must travel with it, and the post-deployment steps (agent user, permissions, activation) on a go-live checklist.
- Deploying prompt templates from sandbox to production~7 minHow prompt templates move between orgs with change sets, Salesforce CLI and packages: deploying referenced flows, what happens to versions and activation, the two Einstein Setup deployment settings, custom model names, and why templates can't be deleted by deployment.
Governance and Observability
10% of the exam- Agent monitoring, session tracing and alerts~7 minWhere agent monitoring sits in the agent development lifecycle, how to turn on Agentforce Session Tracing and install Agent Analytics so data lands in Data 360, how to read one conversation's trace, and how to set and investigate agent health alerts.
- Agent Analytics metrics and optimization~6 minWhat each Agent Analytics metric measures (deflection, escalation, abandonment, engagement, success, task resolution), how LLM-judged quality scores point to retrieval or response problems, how intents are extracted and clustered, and the loop that turns a weak subagent or intent cluster into a measured fix.
Multi-Agent Orchestration
5% of the exam- When to use a multi-agent architecture~7 minWhen one agent should become several and when it shouldn't: the reliability and control benefits of narrow specialist agents, the supervisor or orchestrator front door, Greeter, Operator and Orchestrator patterns, the four orchestration archetypes (SOMA, MOMA, Salesforce-led and MuleSoft-led multi-vendor), and what Agentforce Multi-Agent Orchestration supports today.
- MCP and A2A open agent protocols~7 minWhy Agentforce uses open standards: MCP connects an agent to external tools and data through a host, client and server, with registered and allowlisted server tools becoming agent actions; A2A delegates work to another reasoning agent across orgs and vendors. Includes how admins govern MCP servers and the tool-versus-agent rule for choosing.
Common questions about the Agentforce Specialist study guide
Is the Agentforce Specialist study guide free?
Partly. Every unit's summary and its "What the exam expects", "Key facts" and "Check yourself" sections are free to read without an account. The "Core concepts", "Exam traps" and "Worked scenario" sections, reading progress and unlimited timed mock exams come with the Agentforce Specialist pack (USD $19).
How current is the Agentforce Specialist study guide?
It is written to the current official Salesforce Agentforce Specialist exam outline, and each unit is checked against Salesforce documentation before it is published. The most recent update was October 15, 2026, and every unit shows its own last-updated date.
How does the study guide map to the official Agentforce Specialist exam guide?
There is one unit per topic in the official exam guide, grouped under its 6 weighted sections: Prompt Engineering (20%), Data 360 Fundamentals (20%), AI Agents (35%), Testing, Deployment, and Maintenance (10%), Governance and Observability (10%), Multi-Agent Orchestration (5%). Each unit quotes the exam objective it covers word for word.
How do the practice questions link to the study guide?
Each unit lists the reviewed practice questions that test it (305 questions are linked across the guide), so you can read a topic and practice it in one click. After a practice exam, every missed question links back to the unit that teaches it.