9.6 Agentforce Risk, Troubleshooting, and Governance

Salesforce Admin Study Guide Exam

Track your progress as you read

Finished sections get a check mark.

1 Chapter 1: Platform Administrator Orientation and Official Source Control

1.1 Credential Name, Current Outline, and Search Language 1.2 Exam Format, Fee, Score, and Reference Rules 1.3 Trailhead, Trailhead Academy, and Pearson VUE Delivery Workflow 1.4 Domain Weights, Agentforce, and Study Prioritization 1.5 Maintenance Modules, Credential Ethics, and Exam Security 1.6 Developer Edition, Trailhead Playground, and Practice Bank Workflow

2 Chapter 2: Configuration, Setup, User Management, and Org Controls

2.1 Company Information, Fiscal Year, Business Hours, and Locale 2.2 User Records, Licenses, Profiles, Permission Sets, and Login Access 2.3 Login Security, Session Settings, and Identity Basics 2.4 Setup Navigation, Release Readiness, and Change Control 2.5 UI Settings, Home Pages, Apps, and Navigation 2.6 Configuration and Setup Case Lab

3 Chapter 3: Security, Access, Identity, and Sharing Model

3.1 Org-Wide Defaults, Roles, Sharing Rules, and Teams 3.2 Profiles, Permission Sets, Permission Set Groups, and Muting 3.3 Field-Level Security, Page Layouts, and Dynamic Forms 3.4 Record Access Troubleshooting and Login-As Workflow 3.5 Identity, Login History, MFA, and Session Risk 3.6 Data Visibility, Reporting, and Analytics Security 3.7 Security and Access Case Lab

4 Chapter 4: Object Manager, Lightning App Builder, and Data Modeling

4.1 Standard Objects, Custom Objects, and Schema Fit 4.2 Fields, Formulas, Rollups, Validation, and Requiredness 4.3 Lookup, Master-Detail, and Junction Relationships 4.4 Record Types, Business Processes, Page Layouts, and Lightning Pages 4.5 Dynamic Forms, Dynamic Actions, and User Experience 4.6 Schema Change Risk, Data Quality, and Deployment Readiness 4.7 Object Modeling Case Lab

5 Chapter 5: Sales, Marketing, Service, and Support Applications

5.1 Lead, Campaign, Account, Contact, and Opportunity Flow 5.2 Sales Processes, Path, Products, Price Books, and Forecasting 5.3 Case Management, Queues, Assignment, Auto-Response, and Escalation 5.4 Knowledge, Entitlements, Service Console, and Support Productivity 5.5 Sales-Service Data Quality and Automation Boundaries 5.6 Application Configuration Troubleshooting 5.7 Sales and Service Applications Case Lab

6 Chapter 6: Data and Analytics Management

6.1 Data Import Wizard, Data Loader, Export, and Backup 6.2 Duplicate Management, Validation, and Data Quality Controls 6.3 Data Transfer, Ownership, Mass Update, and Mass Delete 6.4 Report Types, Filters, Summary Formulas, and Buckets 6.5 Dashboards, Components, Subscriptions, and Sharing Impact 6.6 Analytics Security, Visibility, and Executive Reporting 6.7 Data and Analytics Case Lab

7 Chapter 7: Automation, Flow, Approval, and Operational Processes

7.1 Automation Tool Selection and Order of Execution 7.2 Flow Builder: Screen, Record-Triggered, Scheduled, and Autolaunched Flows 7.3 Approval Processes, Email Alerts, and Notification Patterns 7.4 Validation Rules, Assignment Rules, and Escalation Rules 7.5 Automation Testing, Debugging, and Governance 7.6 Automation Performance, Recursion, and Maintainability 7.7 Automation Case Lab

8 Chapter 8: Productivity, Collaboration, Mobile, and AppExchange

8.1 Activities, Tasks, Events, Calendar, and Email Productivity 8.2 Chatter, Feeds, Groups, Notifications, and Collaboration Governance 8.3 Salesforce Mobile Navigation, Actions, and Page Design 8.4 List Views, Kanban, Path, and User Adoption Tools 8.5 AppExchange Evaluation, Installation, Permissions, and Risk 8.6 Productivity and Collaboration Case Lab

9 Chapter 9: Agentforce for Platform Administrators

9.1 Agentforce Use Cases, Admin Boundaries, and Fit 9.2 Trusted AI, Security, Data Access, and Permission Implications 9.3 Agent Builder, Prompt Builder, and Grounding Concepts 9.4 Testing, Deployment, Monitoring, and Feedback Workflows 9.5 Agentforce Service, Sales, and Internal Productivity Scenarios 9.6 Agentforce Risk, Troubleshooting, and Governance 9.7 Agentforce Case Lab

10 Chapter 10: Integrated Salesforce Admin Scenario Labs

10.1 New Sales Org Setup Integrated Lab 10.2 Service Team Case Management Integrated Lab 10.3 Data Quality and Analytics Integrated Lab 10.4 Security, Access, and Reporting Integrated Lab 10.5 Flow Automation and Change Control Integrated Lab 10.6 Agentforce and Adoption Integrated Lab 10.7 Full Admin Simulation Strategy

11 Chapter 11: Final Review, Maintenance, and Test-Day Strategy

11.1 Final 30-Day Salesforce Admin Review Plan 11.2 Trailhead Playground Hands-On Drill Plan 11.3 105-Minute Exam Timing, Flagging, and Review Workflow 11.4 Test-Day Checklist: Online or Test Center 11.5 Post-Exam Maintenance Modules and Next Certifications 11.6 Salesforce Admin Final Mixed Review

9.6 Agentforce Risk, Troubleshooting, and Governance

Key Takeaways Explain This Topic AI

Agentforce risks include data leakage, hallucinated answers, stale grounding, overbroad actions, prompt injection, poor escalation, and unclear ownership.

Troubleshooting should start with the layer most likely to fail: feature enablement, license, app visibility, permission assignment, grounding source, action configuration, channel, or data quality.

Governance assigns owners for use cases, source content, permissions, testing, feedback, monitoring, retention, and change approval.

Admins should document unsupported requests and refusal behavior so users know when the agent should stop or escalate.

Last updated: May 2026

Risks admins must recognize

Agentforce risk is not limited to dramatic AI failures. Most admin risks are ordinary platform risks expressed through a conversational interface. The agent might retrieve the wrong record, summarize stale data, reveal a hidden field, run an action with unexpected side effects, cite an outdated article, fail to transfer a customer, or make a confident statement when the source is weak. Good governance keeps these risks visible and manageable.

The first risk category is data exposure. If grounding includes internal files, old drafts, private notes, or records with weak sharing, the agent can produce answers that users should not receive. The second category is inaccurate output. A fluent answer can still be wrong if the source is stale, incomplete, or ambiguous. The third category is action risk. Creating, updating, routing, or escalating records can trigger validation rules, flows, approvals, assignment rules, notifications, and integrations.

Symptom Likely cause Admin investigation

User cannot access Agentforce Studio Feature, app visibility, profile, tab, or permission issue Check Agentforce enablement, app launcher visibility, assigned apps, tab settings, and Manage AI Agents.

Agent cannot answer from knowledge Article not published, wrong channel, data category restriction, or grounding issue Verify publication, visibility, categories, language, and source connection.

Agent exposes too much information Overbroad grounding or permissions Test as restricted users and remove sensitive sources.

Action fails during conversation Validation, required field, flow fault, permission, or sharing issue Review action context, error logs, field requirements, and automation.

Agent gives poor answers Weak instructions, stale source, ambiguous prompts, or data quality Review transcripts, update content, narrow scope, and retest.

Transfers do not reach reps Channel or routing configuration issue Check queue, Omni-Channel, chat, flow, or handoff settings.

Troubleshooting should be layered. Start with availability: is the feature enabled, is the edition or add-on present, and does the user have the required permission set? Next check navigation: is Agentforce Studio visible in the App Launcher, is the app assigned, and are tabs visible? Then check the agent configuration: is the right agent active, is the channel connected, are subagents or topics enabled, and are instructions saved? Finally check data, actions, and runtime context.

When an answer is wrong, do not edit instructions blindly. First identify whether the answer came from an approved source, generic reasoning, a stale article, missing record data, or a misunderstood request. If the source is wrong, fix the source. If the source is missing, add approved grounding. If the request is unsupported, improve refusal or escalation behavior. If the prompt is ambiguous, add examples or narrower instructions. Retest the exact transcript and related edge cases.

Governance model checklist:

Use-case owner approves what the agent may and may not do.

Content owner approves knowledge, documents, and data libraries used for grounding.

Security owner approves permissions, sharing, channel exposure, and retention.

Admin owner manages configuration, activation, access assignments, and monitoring.

Developer owner supports Apex, external actions, integrations, and complex flow errors.

Support owner triages user issues, feedback, and incident reports.

Change owner controls versioning, release notes, rollback, and deactivation.

Prompt injection and instruction conflicts deserve attention. Users may ask the agent to ignore instructions, reveal internal rules, provide restricted data, or perform unsupported actions. An admin cannot solve this only with a longer instruction paragraph. The agent should have clear refusal behavior, restricted grounding, limited actions, and security tests that include adversarial prompts. Sensitive operations should still require permission checks and approvals outside the conversation.

Audit and retention are part of governance. Conversation logs, feedback, generated outputs, and action records can contain sensitive data. The admin should know who can view monitoring data, whether it is stored in Data Cloud or another supported location, how long it is retained, and how it is used for improvement. Monitoring access should be limited to people who need it for support, compliance, or product ownership.

Study trap: do not troubleshoot Agentforce as one black box. Break the problem into feature enablement, app access, user permission, data source, grounding, action, automation, sharing, channel, and monitoring layers. This approach leads to better answers in scenarios and better support in a real org.