zuiqiang-shifu-topic-title-lab

Generates click-worthy external titles from existing content directions without summarizing the content.

11|1|Updated Aug 3, 2026
One-click install
npx skills add https://github.com/2282199198/zuiqiang-shifu --skill zuiqiang-shifu-topic-title-lab-2282199198
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: zuiqiang-shifu-topic-title-lab
Source: https://github.com/2282199198/zuiqiang-shifu/tree/main/topic-title-lab
Command: npx skills add https://github.com/2282199198/zuiqiang-shifu --skill zuiqiang-shifu-topic-title-lab-2282199198

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Creators who already have a content direction often get titles that simply summarize the whole piece, leaking the answer, method, and conclusion before the viewer clicks. This Skill separates the topic core, the external title, and the content promise so titles create curiosity, identification, or contrast without spoiling the body. ## Core Features & Use Cases - Three-layer separation: Distinguishes the internal topic core, the audience-facing title, and the content promise the body must fulfill. - Mode selection: Chooses among hook titles (curiosity, identification, contrast, benefit), story titles, and search/utility titles based on the task rather than forcing one formula. - Fact whitelist and leakage checks: Every concrete fact, number, or metaphor in a title must trace back to user-provided material, and unverified hypotheses cannot be presented as settled answers. - Use Case: A short-video creator has a confirmed topic about workplace productivity and asks for five titles only. The Skill freezes the content promise, builds a fact whitelist, generates five candidates across at least three different drivers, removes summary information, and outputs numbered titles without any body content. ## Quick Start Tell the AI you already have a content direction and want only external titles, then provide your topic, target audience, and desired number of title candidates.

Frequently Asked Questions about zuiqiang-shifu-topic-title-lab

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write video titles that don't give away the content?

Freeze the content promise first, then pick one driver such as curiosity, identification, contrast, or benefit. Delete supporting facts, reasons, methods, and conclusions from the title while keeping enough relevance that the target viewer knows it concerns them.

What is the difference between a topic and a title for short videos?

The topic core is the internal question or mechanism the creator will explain, while the title is the external hook that makes the target audience stop and click. The title shares the audience and core problem with the topic but omits the full story, causes, steps, and conclusions.

When should I use a story-style title instead of a hook title?

Use a story title only when the user explicitly requests case-based or character-driven framing and provides real material to support it. Hook titles are the default for knowledge, opinion, and workplace content, and facts the user did not provide must never be invented.

Can this Skill find new content topics from scratch?

No, it only handles title packaging after a content direction, core question, or internal shortlist already exists. When there is no direction at all, the request routes back to the upstream topic-engine module for topic discovery.

Why do AI-generated titles sometimes contain facts not in my material?

Models often expand quantities, scenes, or degrees into stronger claims that sound plausible. This Skill prevents that with a fact whitelist requiring every concrete noun, number, and metaphor in a title to trace back word-for-word to user-provided material.