What problem does it solve?
Ghostwriting for founders, colleagues, or yourself usually degrades into generic AI-sounding text because the writer has no disciplined record of how the person actually writes. This Skill grounds every draft in a validated voice profile stored in the GBrain knowledge base, so tweets, posts, launch copy, and emails match the subject's real fingerprint instead of an improvised imitation.
Core Features & Use Cases
- Profile-gated drafting: Loads the subject's voice profile at people/<slug>-voice plus first-party context from the brain, and hard-stops if no validated profile exists rather than freehanding a voice.
- Voice profile builder: Walks through consent recording, first-party corpus gathering (20+ samples over 6+ months), fingerprint extraction across six categories, and a blind validation test before a profile unlocks drafting.
- Self-check before output: Produces 2-3 draft options and scores each against the profile's tells, length band, register purity, and banned boilerplate before showing anything; never auto-posts.
- Use Case: Ask for a launch tweet in a founder's voice; the Skill loads their validated profile, drafts three options in the correct register, and reports a self-check verdict for each.
Quick Start
Ask the agent to draft a tweet as a specific person about a topic, and it will load their validated voice profile from the brain and return checked draft options.
