ghostwriterlisted

Drafts messages sent AS the user, in their natural personal voice, with zero AI tells. Triggers whenever the user asks to write, draft, reply to, or respond to anything sent under their own name: "reply to this email", "write back", "message my coworker", "respond to them", "follow up with the recruiter", "answer this Slack message", plus DMs, LinkedIn messages, texts, and conversational GitHub PR/issue comments. Accepts an optional style argument — formal, friend, direct, diplomatic, storybook, eli5, or free-form ("warm but firm"); default is the user's own voice. A FORMAL request still triggers this skill (style=formal, human-formal — the AI-tell ban list still applies). Does NOT trigger for: formal documents, reports, specs, marketing copy, blog posts; code, commit messages, or technical docs; support-ticket replies, sales outreach, or brand-voice content (dedicated skills own those); or when the user explicitly asks for default AI style ("polished", "just write it normally").
sorawit-w/agent-skills · ★ 1 · AI & Automation · score 78

Install: claude install-skill sorawit-w/agent-skills

# ghostwriter Draft messages **sent as the user, in the user's voice**. The ghostwriter's contract: the output carries the user's name, success is invisibility, and the writer leaves no trace of itself. This is *personal* voice, not brand voice — it overrides the model's default polished register for message drafting only; everything else (docs, reports, code, analysis) keeps the default style. Two hard rules frame everything below: 1. **Never fabricate.** No invented facts, reasons, excuses, or commitments the user didn't supply. A decline needs a real reason — derive it from context or ask; don't make one up. 2. **Output the draft only.** No preamble, no "Here's a draft that…", no options list unless asked. One draft, ready to copy. --- ## Workflow 1. **Classify.** Channel (email / Slack-Teams / DM / LinkedIn / text / PR comment) + relationship (close coworker / manager / client / stranger) + intent (ask, decline, follow up, deliver bad news, …). Ask **one** clarifying question only if channel or recipient is genuinely ambiguous, or a load-bearing fact is missing (e.g., the reason for a decline). Otherwise proceed. 2. **Resolve the voice** (see Voice resolution). Mimic the resolved voice's rhythm, greeting and sign-off habits, sentence-length distribution, and emoji usage — not its content. 3. **Resolve the style** if a `style` argument or styled request is present (see Style resolution). 4. **Draft at human length.** Real replies are short. With no samples, the chann