episodelisted

Split one source post into tier-sized drafts (teaser, main social post, longread, technical dev-log) with cross-links downward and a canonical link back to the repository. Draft-first: publishes nothing itself. Triggers: "/episode", "/adapt", "adapt this post", "make an episode".
tonydzi/second-brain-starter-kit · ★ 6 · AI & Automation · score 78

Install: claude install-skill tonydzi/second-brain-starter-kit

EPISODE_ADAPTER (stage S5 of content factory v2 — the contract [[decision-content-pipeline-reality-show]] §7.2) GOAL: ONE source post → an "episode" = one draft PER (tier × platform) from §7.2, with cross-links downward and canonical→GitHub baked in. Draft-first: we write drafts and publish NOTHING. THE §7.2 CONTRACT (implemented 1:1): ``` tier length platforms (phase 1) language voice/model teaser 240-370 (HARD) RU channel · X (EN) RU TG/EN X authorial · Opus medium 800-2500 Facebook (RU) RU authorial diary · Opus · LINK IN THE 1ST COMMENT (not in the body) + a question longread 4000+ narrative RU channel·VC.ru·Habr(RU) · GitHub(EN) RU and EN authorial reality-show · Opus dev-log a dry log GitHub (EN, canonical) EN technical, impersonal · Sonnet scaffold + Opus for cohesion ``` - The owner's voice (`authorial`) = **Opus ONLY**. The dev-log is not authorial — Sonnet assembles it, Opus stitches it together. - **canonical = the GitHub markdown** on every platform (a single source of truth for AI). Repo: `github.com/tonydzi/clawrush`. - **NO copy-paste** — every file is a NATIVE rewrite for its platform. - Cross-links downward: teaser→longread; medium→longread (in the 1st comment); longread ←teaser →dev-log; dev-log→longread + an anchor to our Deep Research. - THE PLATFORM DECIDES THE LANGUAGE (there is no bilingual pair on every tier). - Th