ghostwriter desk persona

The problem

Most AI writing tools fail because they default to a generic "assistant" tone and invent facts to fill gaps. This creates drafts that sound like AI and require extensive manual cleanup to match a client's actual voice.

What it does

• Builds a structured voice profile from source material before drafting to ensure tonal consistency.
• Maps every factual claim, anecdote, and opinion back to original client sources or transcripts.
• Flags unsourced information as NEEDS SOURCE instead of hallucinating details.
• Orchestrates revisions in convergent stages, separating structural changes from final line edits.
• Identifies and removes "written" prose that a client would never actually say out loud.

Why this beats prompting it yourself

Generic prompts often result in "wordy" or "helpful" AI personas that drift into flowery language. This skill enforces a project-managed editorial process that treats revision as a contract and strictly adheres to evidence-based drafting, preventing the common "vibes-based" output of standard LLMs.

Use cases

• Ghostwriting executive newsletters and LinkedIn thought leadership from call transcripts.
• Drafting founder bylines that maintain specific industry jargon and personal anecdotes.
• Converting raw interview notes into structured book chapters or long-form articles.
• Managing multi-round revision cycles without losing the original voice or scope.

Known limitations

Does not handle academic submissions, legal documents, or medical texts requiring licensure. Requires source material or a voice profile to be provided before drafting can begin.