Documentation
The operating manual.
How Siren is configured, how a prompt becomes an asset, and how posting is recorded. Begin with Getting started if this is your first workspace.
Onboarding
Configuration
Brand DNA. What each field does
Colors, fonts, archetype, banned words, required disclaimers, and how server-locking keeps them from drifting.
Mascot. One character, in every post
How Siren makes your mascot, what you get with it, and how it shows up in your stills and videos.
Workspaces. Multi-brand isolation and seat management
How workspaces separate Brand DNA, assets, and permissions. When to use more than one.
Generating
The prompt. Writing one Siren can render
What a prompt needs: topic, angle, optional format hint. What to avoid. Examples by asset type.
Asset types and channels. When to pick which
Cards, videos, slideshows, carousels, infographics, OG images. What each is for, how they render, and which channels they suit.
Variation axes. Composition, mood, layout, more
How Siren produces distinct outputs from the same template, and how the Tweaks panel changes an axis after a run.
Posting
OAuth dispatch and audit trails
How Siren dispatches posts to connected platforms, what the audit trail captures, and how to trace a dispatch if something goes wrong.
Connect an agent over MCP
Add Siren to Claude Code, Claude, Cursor, ChatGPT, or Grok. One approve, then the agent can plan, render, and post.
Company