Visual Intelligence System
Drop images, get a per-platform strategy.
The connective tissue between asset generation, composition, and ship. Three asset tiers (NB2, Higgsfield, HyperFrames), nine platform personas, one drop-images workflow. Substrate at lib/visual-intelligence/, runtime at /visual-strategy + visual-intelligence skill + visual-intelligence-orchestrator agent.
LinkedIn (AI Architect), YouTube long-form (the Builder), Shorts + TikTok (the Creator), Instagram (the Aesthete), X (the Thinker), Threads (Conversation-starter), Bluesky (Live-thinker), Spotify (the Producer). One persona per profile.
- Per-batch strategy
- Per-image analysis
- Gap-fill briefs
- Persona matrix
/visual-strategy/vis-search/vis-audit/infogeniusVideo capture to published short with AEO schema.
Talking-head capture to publish-ready video in one shot.
Plan, deliver, and amplify live workshops end-to-end.
Book intelligence — quotes, chapters, continue-reading, videos.
The kata for personal capital stewardship.
Try Visual Intelligence System yourself
The full stack is being packaged as an installable template. One command scaffolds a working version of this module in your own project.
MIT license · portable across Claude Code, ChatGPT, Codex, Gemini CLI