Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Short-Drama Know-How Learning & Promotion Governance

A maintainer-only workflow for qualitatively studying private short-drama project chains and promoting findings into public references only after de-identification, blind eval, and independent review.

AutomationAdvanced★ 1,797⑂ 433AI score 6/10Last updated: Sep 11, 2026

What it does

  • Guides the agent to read maintainer-authorized, read-only private text sources (creative briefs, story engines, episode scripts, character/scene/prop continuity ledgers, image & video prompts, review and revision notes) as complete project chains rather than keyword mining.
  • Separates observation cards (direct observation vs. agent interpretation vs. unknown) from decision cards (whether a candidate was formed and why scope was narrowed).
  • Builds a genre × dramatic-mechanism coverage matrix that tracks supporting cards, counterexamples, boundaries, and the next sample to read — never frequencies or success rates.
  • Enforces de-identification and de-copy: strip names, exact lines and rare combinations, then rebuild examples in a brand-new genre and wording.
  • Requires a fresh-agent blind forward eval plus an independent reviewer before any minimal change is promoted to the public skill suite, with explicit narrow/replace/retire rules afterward.

Who it's for

  • Maintainers of a short-drama or scripted-content skill library
  • Teams that must convert confidential client material into safely publishable guidance
  • Researchers who prefer qualitative validation over frequency statistics
  • ⚠️ Not intended for everyday scriptwriting, public runtime data access, or media generation.

Example uses

  1. Sample 12 romance and revenge projects from an authorized archive, card the episode-ending hook mechanisms including failure cases, and locate empty matrix cells for the next reading round.
  2. Take a tentative rule like "a twist must pay off within three episodes," hunt counterexamples, and rewrite it as "valid only under these audience promises, character states, and production constraints."
  3. Hand an anonymized candidate to a fresh agent on synthetic tasks without revealing expected answers, compare against the public baseline, and promote only after an independent reviewer clears leakage and mechanization risks.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/zenstory-ai/drama-skills/HEAD/maintainers/skills/short-drama-knowhow/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the maintainers/skills/short-drama-knowhow folder from the GitHub repo zenstory-ai/drama-skills into my ~/.claude/skills/short-drama-knowhow/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/zenstory-ai/drama-skills.git && mkdir -p ~/.claude/skills && cp -r drama-skills/maintainers/skills/short-drama-knowhow ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Clone the repository: git clone https://github.com/zenstory-ai/drama-skills.git
  3. Create the skills folder if needed: mkdir -p ~/.claude/skills
  4. Copy the skill: cp -r drama-skills/maintainers/skills/short-drama-knowhow ~/.claude/skills/
  5. Verify that references/ contains cards-and-coverage.md, synthesis-and-promotion.md, blind-forward-eval.md, promotion-ledger.md, and prompt-production-learning.md.
  6. Restart Claude Code and invoke it explicitly, e.g. "Start short-drama know-how learning from the authorized read-only source."
  7. Before running, write down this round's allowed reading scope, forbidden content, output location, and destruction requirements.
View source on GitHub ↗License: MIT