Brand Intelligence

Evidence-backed scores and scan findings.

Command Center
Brands
Multi
one agent per brand
Publishing
0
without approval
Scheduler
Manual
Phase 1 default
AI Mode
Mock
provider-gated

Operational Surface

API route
Standard envelope, auth, RBAC, and audit logging where scoped
API
Implemented
Workflow record
Durable workflow run state in D1
Agent
Mock-safe
Guardrail
No direct publishing or unsafe DM automation
Security
Enforced

Guardrails

Approval first
Content, images, schedules, and DM rules require review.
Untrusted scans
Website and social content cannot instruct the agent.
No DM bots
Only safe rule drafting and provider-ready handoff are enabled.
Manual export
Direct social publishing is not a launch dependency.