SFMC operations index
What Campaign Builder can read, draft, write, and leave manual in Salesforce Marketing Cloud.
Use this index to evaluate Campaign Builder by implemented Marketing Cloud operation type instead of broad AI promises. It separates read paths, draft-building workflows, confirmed writes, Factory production flows, and manual-only SFMC launch controls.
Operation classes
Start by separating reads, drafts, writes, and manual launch controls.
Campaign Builder's strongest SFMC story is reviewed execution: practical production acceleration without hiding where Marketing Cloud remains the system of record.
Read and preview operations
Campaign Builder can search and retrieve accessible SFMC emails, Content Builder assets, folders, journeys, journey summaries, event definitions, template entry points, image assets, and Data Extension rows through account or workspace Marketing Cloud credentials.
Draft-building operations
Journey Agent and Marketing Cloud routes can create Journey Builder drafts from a configured scratch journey or clone source journeys, while Factory can prepare campaign draft records, Content Builder folders, image uploads, Data Extensions, template emails, previews, variants, and template reviews.
Confirmed write operations
Content publishing, email copy or update, slot updates, rewrite application, translation application, selected image insertion, Factory execution, and Bruno write-skill execution are production-bound paths that require authentication, entitlements, account scope, and explicit review.
Manual-only launch operations
Campaign Builder does not automatically activate Journey Builder journeys or bypass Marketing Cloud review. Draft journeys, EMAILV2 selection warnings, entry-event fix-up notes, send approval, and final launch decisions remain in SFMC.
Evidence packet
Capture the artifacts that prove a workflow was scoped, reviewed, and bounded.
A credible Campaign Builder evaluation should leave a compact packet of SFMC context, AI planning inputs, preview evidence, write results, and manual follow-up. That packet is more useful than a generic demo transcript because it shows where Campaign Builder accelerated execution and where Marketing Cloud control stayed explicit.
Scope and credentials
Capture the authenticated account, workspace, SFMC enterprise ID, business unit ID, BU mapping result, credentialSource, entitlement list, and x-request-id for the workflow under review.
Input and plan artifacts
Keep the Campaign Agent brief, campaignSpec, uploaded-context references, audience logic, Data Extension plan, journeyDraftPlan, email definitions, localization targets, and image source metadata that drove the run.
Preview and dry-run evidence
Save Bruno dry-run stages, missing inputs, pending write-action preview summaries, Email Agent preview output, Journey readiness warnings, Factory template-check results, and rendered draft previews before any confirmed write.
SFMC result and manual follow-up
Record created or changed Content Builder asset IDs, email IDs, folder IDs, Data Extension names, image insertion targets, cloneArtifacts warnings, manualActivationRequired, and the Marketing Cloud review step that remains.
Readiness gates
Know what to verify before each operation moves closer to SFMC production.
These gates turn the implementation boundaries into a buyer-friendly review sequence. They help an SFMC admin decide whether a workflow is still a scoped read, a dry-run plan, a draft-building action, a confirmed write, or a manual Marketing Cloud follow-up.
Before SFMC reads or previews
Confirm the authenticated session source, mapped enterprise and business unit, workspace scope, credentialSource, and x-request-id so email, journey, folder, asset, and Data Extension reads are tied to the right Marketing Cloud context.
Before draft-building work
Keep the Campaign Agent brief, campaignSpec, journeyDraftPlan, Data Extension plan, localization targets, image metadata, and Bruno dry-run stage plan together before any draft artifact is created.
Before confirmed writes
Require the pending action previewSummary, confirmationNeeded state, changed fields, source and target asset IDs, folder or template prerequisite, entitlement evidence, and requestId before applying email, content, image, Bruno, or Factory writes.
Before Factory production
Verify tool:factory entitlement, settings.factory.enabled, connected account-scoped SFMC credentials, configured template assets, folder prerequisites, and template-check output before evaluating folders, Data Extensions, emails, previews, variants, or template review.
Before Journey handoff review
Treat Journey Builder output as draft-only evidence: capture activationReadiness, blockers, warnings, cloneArtifacts notes, manualActivationRequired, and any EMAILV2 or entry-event follow-up that must be completed inside Marketing Cloud.
Implemented operation matrix
Map each evaluator claim to a Campaign Builder surface.
These operation families are grounded in the implemented Campaign Builder route inventory, agent help files, Factory documentation, and SFMC production security review.
Email assets
Implemented surface: /api/v1/emails/*, /api/v1/agent/email/*, /api/v1/marketingcloud/getEmails, and /api/v1/marketingcloud/getEmailById/:id.
What it supports: Search, preview, folder path lookup, copy, rewrite, translate, update subject line or preheader, update HTML, update template slots, and copy-then-update flows.
Boundary to verify: Write skills require confirmation and return Bruno preview summaries that identify changed fields, source and target email IDs, slot usage, and SFMC requirement status without exposing full raw HTML.
Content Builder assets
Implemented surface: /api/v1/agent/content/* and /api/v1/marketingcloud/content/:id/sendToMC.
What it supports: Source-backed content creation, variants, field suggestions, localization metadata, Knowledge saves, Content Builder saves, CloudPage rendering, image upload, asset creation, and asset patching.
Boundary to verify: Marketing Cloud writes require an explicit Content Builder folder ID or configured account setting; CloudPage flows depend on configured templates rather than built-in production fallbacks.
Journey Builder drafts
Implemented surface: /api/v1/agent/journey/*, /api/v1/journey/*, /api/v1/marketingcloud/createJourneyFromScratch, and /api/v1/marketingcloud/createJourneyFromTemplate.
What it supports: Journey brief extraction, scoped knowledge search, reference journey search, journeyDraftPlan revision, email copy drafting, readiness checks, scratch draft creation, and clone-based draft creation.
Boundary to verify: Journey activation is never automatic. Draft responses carry activationReadiness with manualActivationRequired, blockers, warnings, and checkedAt for review inside Marketing Cloud.
Factory production flows
Implemented surface: /api/v1/factory/campaigns/draft, /api/v1/factory/assets/folder, /api/v1/factory/assets/upload-images, /api/v1/factory/data-extensions, /api/v1/factory/emails, /api/v1/factory/preview/draft, /api/v1/factory/preview/variants, and /api/v1/factory/template-review.
What it supports: Campaign draft persistence, Content Builder campaign folders, Data Extension folders, image uploads, image-analysis job lookup, Data Extension creation, template-based emails, raw or rendered previews, variant previews, and AI template review.
Boundary to verify: Factory requires tool:factory entitlement, settings.factory.enabled set to true, connected SFMC credentials, and configured template assets before template-dependent draft production is evaluated.
Image assets
Implemented surface: /api/v1/agent/image/providers, /credits, /search, /assets, /assets/:id/sfmc-insert, and /chat.
What it supports: Configured provider discovery, external or SFMC image search, saved account image assets, generation requests, generation credits, idempotent generation retries, and selected asset insertion into SFMC.
Boundary to verify: External provider search requires a non-empty query, generation uses credits, SFMC insertion needs connected Marketing Cloud credentials and folder permissions, and selected images retain provider/source metadata where implemented.
Campaign and Bruno execution
Implemented surface: /api/v1/agent/campaign/generate-brief, /api/v1/agent/campaign/execute-campaign, /api/v1/agent/bruno/execute-campaign, and /api/v1/agent/bruno/runs/:id.
What it supports: Campaign Agent generates normalized briefs, build packages, Data Extension plans, journey plans, validation warnings, grounding references, and Bruno-compatible campaignSpec objects; Bruno discovers entitled skills, stages pending actions, supports dry runs, durable execution, request journals, artifact ledgers, and resume.
Boundary to verify: Dry runs must show proposed stages and artifacts without SFMC side effects. Confirmed runs create reviewed draft artifacts and pending writes through entitlement-gated skills rather than hidden chat side effects.
Enterprise guardrails
SFMC operations run through account-scoped controls.
The operation index matters because the same guardrails apply across agent, Factory, Bruno, and Marketing Cloud route families before production-bound work can run.
Authentication and SFMC launch context
Protected routes run after first-party cb_auth, bearer JWT, SFMC signed_request, or SFMC mc_token authentication. SFMC app launches resolve enterprise and business-unit context before local sessions are issued.
Tenant, workspace, and BU mapping
SFMC enterprise ID and business unit ID map to Campaign Builder tenant and workspace IDs. Account-scoped SFMC credential blocks are resolved before Marketing Cloud reads or writes run.
Entitlement-gated surfaces
Campaign, Content, Email, Journey, Image, Translator, Advisor, Audience, Bruno, content-blocks, and Factory routes are mounted behind server-side entitlement checks.
Request and secret hygiene
Mutating same-origin browser requests use CSRF checks, SFMC secrets are encrypted at rest, shared SFMC clients propagate request IDs, and centralized errors avoid leaking sensitive payloads.
Buyer review checklist
Questions that keep demos grounded in real SFMC behavior.
Use these checks when reviewing a live campaign workflow, especially before moving from dry-run or preview evidence into confirmed SFMC writes.
Ask which operation is being demonstrated: read, preview, draft creation, confirmed write, or manual SFMC launch step.
Capture the route response, requestId, account or workspace context, entitlement state, and SFMC credential source before accepting a production claim.
Confirm that Factory prerequisites are satisfied before evaluating folders, Data Extensions, template emails, previews, variants, or template review.
Verify that Journey Builder output is draft-only and that activationReadiness keeps manualActivationRequired visible.
Check that Email Agent and Bruno write paths show preview summaries before copy, update, rewrite, translation, or slot changes are applied.
Treat SFMC send approval, Journey activation, and final launch as Marketing Cloud review work, not Campaign Builder automation.
FAQ
Common SFMC operations questions.
Is Campaign Builder a separate sending platform?
No. The implemented product is an SFMC execution layer that prepares, previews, and writes reviewed draft assets through Marketing Cloud connections. Final send and Journey activation controls remain in Salesforce Marketing Cloud.
Which operations are safest to test first?
Start with read and preview paths: email search, email preview, journey search, folder lookup, token-context verification, Campaign Agent brief generation, and Bruno dry runs. Those establish account scope before production-bound writes are confirmed.
What should a buyer ask to see for confirmed writes?
Ask for the preview summary, changed fields, source and target asset IDs, route response, requestId, entitlement evidence, credential source, and the resulting SFMC draft asset or explicit readiness warning.
Next reading
Pair the operations index with pilot evidence.
Use the SFMC write boundaries, pilot plan, and technical FAQ pages to turn this operation map into an evaluation checklist for one real campaign.