Setup with Agentforce (Beta)
Setup with Agentforce (Beta) is an experimental Salesforce feature that wraps the Setup interface in a natural-language chat.
Definition
Setup with Agentforce (Beta) is an experimental Salesforce feature that wraps the Setup interface in a natural-language chat. Admins ask the agent to create custom fields, build flows, configure permission sets, and adjust org settings in plain English. The agent reads the request, generates the corresponding metadata change, shows a preview of what will be created or modified, and applies it on confirmation. Work that historically required twenty clicks across multiple Setup pages collapses to one sentence and one approval.
The feature is a Beta as of mid-2026, which means it ships in production orgs behind a feature flag, accepts feedback as the primary signal, and may change behavior between releases. Only a subset of Setup actions is supported, and complex multi-step configurations (deploying a managed package, configuring SSO) still require the traditional UI. The right way to think about Setup with Agentforce is as a fast path for the routine 80 percent of admin work, not a replacement for hands-on Setup expertise.
Why Beta means useful but not load-bearing yet
Where the agent lives and how it is invoked
Setup with Agentforce appears as a chat panel in the Lightning Setup tree once the feature is enabled at the org level. The icon shows at the top of the Setup left navigation. Click it and a sidebar opens where admins type requests. The agent is scoped to Setup operations only; it does not act on user data or records. Behind the scenes the agent calls a curated set of Agent Actions that wrap the Metadata API and the Tooling API. Each action covers one common Setup operation: create custom field, build flow, configure permission set, set field-level security, enable a feature.
What it can do today and what it cannot
Supported in Beta as of mid-2026: custom field creation on standard and custom objects, custom object creation, page layout updates, flow creation from a description, permission set creation and field assignment, profile field-level security changes, validation rule creation, and report folder management. Not yet supported: managed package install, SSO setup, custom domain configuration, named credential setup, Apex code deployment, Lightning App Builder page edits, and anything that requires UI state the agent cannot inspect. The list of supported operations expands per release; check the most recent release notes before relying on a specific capability.
The preview-and-approve safety model
Every change the agent generates is shown as a preview before it applies. The preview lists every metadata file the change will create or modify, the field-by-field diff for existing components, and any dependent changes (a new custom field automatically adds itself to the default page layout, for example). The admin clicks Apply to commit. There is no autonomous execution path in Beta; every change requires explicit approval. The audit trail captures the original request, the generated metadata, the approver, and the timestamp.
How the agent grounds its requests
The agent reads the org's existing metadata at request time. A request to "add a Tier field on Account" prompts the agent to first list existing Account fields and detect that no field named Tier or Tier__c exists, then propose a new field with sensible defaults (Picklist, values High and Medium and Low, page layout assignment to the default layout). The grounding prevents the most common error class: agents confidently creating duplicate fields under slightly different names. The same grounding applies to flow creation; the agent reads existing similar flows and matches their conventions for naming, error handling, and version metadata.
Multi-step requests and conversational refinement
Admins can request multi-step changes ("create a Tier field on Account, then update the Sales Manager permission set to give read-only access to it"). The agent generates the full set of changes and shows one combined preview. Refinement is conversational: after seeing the preview, the admin can ask "change Tier to a multi-select picklist" and the agent regenerates against the existing preview state. Refinement only works inside the same uncommitted preview; once Apply is clicked, the changes are real metadata and further edits require a fresh request.
Permissions, audit, and governance
Setup with Agentforce runs as the calling admin user. Every action the agent applies is subject to the same permissions the admin has when clicking through Setup manually. An admin without Modify All Data cannot use the agent to bypass that restriction. The audit trail writes to the SetupAuditTrail object the same way manual changes do, plus a separate AgentInteraction record that captures the original request and the generated metadata. Compliance teams should expect to review both streams together; a request like "add a custom field" is audit-clean but a request like "make Tier visible to everyone" needs review against the permission-set context.
Where this is going and what to plan for
The Beta is the first public step on a longer roadmap toward conversational metadata operations. Salesforce has signaled that future releases will expand action coverage (Apex deployment, Lightning page edits, managed package management), improve grounding (project-aware suggestions, sandbox-to-production promotion), and integrate with Vibes for code-aware Setup changes. Teams should plan for the agent to become a meaningful productivity layer over the next 12 to 24 months, but not yet plan to retire admin training on the underlying Setup UI. The Beta accelerates routine work; it does not yet replace the depth admins need to debug edge cases.
How to use Setup with Agentforce safely while it is still Beta
The Beta is genuinely useful for routine work and genuinely incomplete. The rule that pays off in practice is: use the agent for the routine 80 percent, keep manual Setup skills sharp for the other 20 percent. Treat every Apply click as a real metadata change that ships through your normal deployment process, including sandbox first.
- Enable Setup with Agentforce on a sandbox first
Setup, Feature Settings, Setup with Agentforce, Enable. Do this in a sandbox so the team can try requests without risk. Production rollout happens after the team is comfortable.
- Try low-risk requests to calibrate expectations
Ask the agent to create a custom field, build a simple validation rule, add a value to a picklist. Observe the preview quality and the agent reasoning. This calibrates what to trust without risk.
- Read every preview before clicking Apply
The preview is the safety mechanism. Reading it carefully catches the edge cases the agent missed (wrong picklist value, wrong page layout, wrong dependent field). Apply without reading is the failure mode that produces incidents.
- Use the agent for multi-step routine work
Requests like "create a Tier field on Account and assign it to the Sales Manager permission set" save the most time. Single-field changes are fast either way; the multi-step requests are where the agent earns its keep.
- Keep manual Setup skills sharp
The Beta does not cover every Setup operation, and complex debugging still requires hands-on UI work. Continue training admins on the underlying Setup tree, not just on agent requests.
- Roll changes to production through your normal deploy path
An agent-generated metadata change is still a metadata change. Push it from sandbox to production through your change set, source-tracked sandbox, or DevOps Center pipeline. The agent is the editor, not the deploy mechanism.
- Track Beta limitations against your roadmap
Watch the release notes for each Salesforce release. Action coverage expands per release, and capabilities that were missing in the Spring may ship in the Summer.
Which orgs (sandbox, production, scratch) have the feature flag on. Default off in production until the admin team is comfortable.
Which permission sets grant access to the Setup chat panel. Defaults to System Administrator; can be relaxed for delegated admins.
Which Setup operations the agent supports in the current release. Tracked in release notes; expands per release.
How much detail the preview shows for dependent changes (default vs verbose). Verbose helps admins still calibrating trust.
Whether AgentInteraction records cross-link with SetupAuditTrail entries for unified compliance review. Enabled by default once the feature is on.
- The Beta does not cover every Setup operation. Action coverage is documented per release; check the release notes before assuming a specific capability is supported.
- The preview is the only safety mechanism between the agent and real metadata. Apply without reading is the most reliable way to produce an incident.
- The agent runs as the calling admin. It does not grant permissions the admin does not have, but it also does not warn when the admin is using their broad permissions on a sensitive object.
- Multi-step refinement only works inside an uncommitted preview. Once Apply is clicked, further changes require a fresh request and a fresh preview.
- Agent-generated changes still need to flow through your normal deploy pipeline. The agent is the editor; the deploy mechanism is unchanged.
Trust & references
Cross-checked against the following references.
- Agentforce overviewSalesforce
- Setup with Agentforce release notesSalesforce
Straight from the source - Salesforce's reference material on Setup with Agentforce (Beta).
- Setup with Agentforce (Beta)Salesforce Help
- Supported Setup ActionsSalesforce Help
- Enable Agentforce in Your OrgSalesforce Help
About the Author
Dipojjal Chakrabarti is a B2C Solution Architect with 29 Salesforce certifications and over 13 years in the Salesforce ecosystem. He runs salesforcedictionary.com to help admins, developers, architects, and cert/interview candidates sharpen their fundamentals. More about Dipojjal.
Test your knowledge
Q1. What does Setup with Agentforce (Beta) need to work effectively?
Q2. What is a key benefit of Setup with Agentforce (Beta) for business users?
Q3. How does the Einstein Trust Layer relate to Setup with Agentforce (Beta)?
Discussion
Loading discussion…