What problem does it solve? Writing messages that sound like a specific person is hard — generic AI output defaults to polished, hedged, corporate prose. This Skill encodes one person's actual writing style (distilled from 11,528 verbatim Slack messages) so drafts match his terse, blunt, Hebrew-English code-switching voice instead of sounding like a content team. ## Core Features & Use Cases - Voice-matched Slack drafting: Produces DMs, channel posts, and replies as 2-4 short burst messages with the correct language choice (Hebrew for Israeli 1:1s, English for channels) and signature patterns like spaced-hyphen connectors and one-word seals. - Situation-specific formulas: Covers critical feedback (gap + expectation), praise (short + specific + !), decisions (בוא נ… + seal), disagreement, and banter, each with real example messages. - Long-form register: Adapts the same voice to blog posts and long internal notes — composed English sentences that keep the dry humor, deadpan sarcasm, and verdict-first directness. - Use Case: You need to tell a report, as Aryeh, that a deadline slipped without warning. The Skill outputs a Hebrew burst: names the gap, states the expectation (פעם הבאה אני מצפה שתרים דגל בזמן), and closes flat — no sandwich, no hedging. ## Quick Start Ask the assistant to draft a Slack message as Aryeh to a specific person about your topic, for example: write a DM to Danny in my voice telling him the migration plan is over-scoped.
