Notifications
Brand Guard posts notices in two places: its own brand-guard channel (for conversational replies) and the channel of whichever agent submitted the draft (when it blocks something). Auto-approvals happen silently; you are notified when action is needed, not when everything is fine.
For history, the in-app Brand Guard workspace inside The Marketing Workspace is the home base. Every check Brand Guard runs lands there with its decision and a brand-fit score band, whether or not a Slack notice went out. The notices below tell you when to act; the workspace is where you look back.
Block Notice (in the Calling Agent's Channel)
When: Brand Guard stops a draft because the score fell below the approval threshold, a prohibited phrase was found, a required disclaimer is missing, or the draft tripped a hard block (an invented or unsupported citation, or an unsupported performance claim). A hard block stops the draft no matter how high it scored.
What it contains:
- A one-line summary: the draft reference, the total score, and the primary reason for the block (for example, a prohibited phrase or a low Voice score).
- A pointer to the full breakdown in the calling agent's card.
Action required: Go to the calling agent's card (Marketing Content, Marketing Social, or Marketing Email) to see the full score breakdown and the specific redlines. From there, revise and resubmit, or use "Publish now" to override the block.
Auto-Approval (Silent)
When: A draft scores at or above the approval threshold with no critical flags.
What it contains: Nothing is posted. The draft moves forward automatically and the decision is recorded in the audit log.
Action required: None. If you want to confirm the score after the fact, ask Brand Guard in the brand-guard channel.
Full Score Breakdown (in the Brand-Guard Channel)
When: You request a conversational review or ask Brand Guard to explain a past decision.
What it contains:
- The total score and a per-dimension breakdown (Voice out of 30, Visual out of 25, Messaging out of 25, Governance out of 20).
- The verdict: auto-approve, review, revise, reject, or gate failed.
- Any flags raised, such as a detected cliche opener, a missing disclaimer, or an off-pillar claim.
- Redlines when the verdict is revise or reject: each redline cites the exact passage, explains the issue, and proposes a specific replacement.
Action required: Depends on the verdict. Auto-approve: no action. Review: your team looks at the advisory redlines and decides whether to apply them. Revise or reject: apply the redlines and resubmit.
Gate Failed Notice
When: Brand Guard cannot complete a check (for example because it could not finish scoring the draft). Rather than let an unchecked draft through, it fails closed and records the outcome as "gate failed."
What it contains:
- A note that the draft was stopped because the check could not finish, not because of a specific brand rule.
- A pointer to the calling agent's card and the in-app workspace, where the gate-failed result is recorded.
Action required: Treat it as a block. Resubmit the draft through the calling agent to run a fresh check. If the same draft keeps failing the check, contact support@agenteous.com.
Missing Brand Configuration Notice
When: Brand Guard cannot find brand rules for the client named in the draft.
What it contains:
- A note that the audit ran against platform defaults rather than that client's specific rules.
- The score and decision reached under those defaults.
- A recommendation to set up the client's brand configuration in the dashboard.
Action required: Set up the client's brand profile in the dashboard so future audits apply the correct rules. The current draft's decision stands as recorded.