pre-event-content

Provide the event name and the Skill fetches the completed research brief from the Notion Content Drafts database, then drafts three hook variants (question, contrarian, stat/story) following the Hook-Context-Insight-CTA structure. You pick a variant, which is then optimized before being written back to Notion.

The Skill generates two variants per speaker or host anchored to different signals: one from their talk abstract and one from their recent external work. Each note is counted against the 200-character free-tier cap, scored against a personalization rubric, and must reach 80 before being presented.

It requires a completed research brief in Notion containing event details, topics, people with connection angles, companies, and verification flags. The Skill reads the brief and reference style files; it does not perform its own research, so a missing or weak brief must be completed first.

Yes. Every LinkedIn post ships with a 3-5 slide carousel brief built from one of four narrative arcs, then auto-rendered through the Gamma MCP integration in 4x5 portrait format. Canva is used only as a fallback when Gamma cannot satisfy a specific brief.

The Skill flags the person as needing more research instead of generating a generic note. It presents the gap and suggests what kind of signal would unlock a note, since a weak connection request burns the single first-touch impression.

Skip optimization when the event is under 12 hours away, when the underlying brief is too weak to support a real insight, or when the first-pass score already exceeds 85 with no flagged dimensions. Optimization only applies to LinkedIn posts, never to DMs or prepared questions.