CommunityCoding & Developmentgithub.com

tangkai1001/title-tag-cover-generator

A two-stage Codex Skill for Chinese content publishing scenarios that generates titles, tags, and cover images from articles, with a human-in-the-loop checkpoint to ensure titles are approved before proceeding to cover generation.

What is title-tag-cover-generator?

The `tangkai1001/title-tag-cover-generator` is a specialized two-stage AI agent skill designed for Chinese content distribution workflows, supporting platforms like Xiaohongshu, WeChat Channels, Douyin, and WeChat Official Accounts. It accepts an article and first generates an article analysis, multi-platform title candidates, an AI-recommended title, tag strategy, and recommended tags based on predefined title strategies and tag rules. The skill then enforces a mandatory pause after Stage 1, requiring human confirmation of the chosen title and tags before proceeding. Only after title approval does Stage 2 begin, generating a cover image prompt and invoking either a mock or an external image generation API to produce the cover, culminating in a complete channel material package JSON. This skill enforces a clear process boundary ('titles and tags first, cover later') to avoid generating useless covers before titles are finalized, and it can be easily ported to other AI agents (e.g., Claude Code, Cursor, Codex) or GitHub repositories. It does not handle automatic publishing, CRM integration, or video generation—focusing solely on the article-to-assets pipeline.

Works with~Claude Code✓Codex CLI~Cursor
npx skills add tangkai1001/title-tag-cover-generator

Installed? Explore more Coding & Development skills: steipete/bluebubbles, steipete/eightctl, steipete/blucli · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

What does title-tag-cover-generator do?

The Title-Tag-Cover Generator is a two-stage Codex Skill tailored for Chinese content publishing scenarios. It receives an article, then generates article analysis, multi-platform title candidates, AI-recommended titles, tag strategies, and recommended tags based on title strategies and tag rules. After Stage 1, it enforces a mandatory pause to wait for human confirmation of the title and tags. Only after the title passes human review does Stage 2 begin, continuing to generate a cover image prompt, calling a mock or external image generation API to produce the cover, and finally outputting a complete channel material package JSON. This Skill is suitable for content distribution preparation on platforms such as Xiaohongshu, WeChat Channels, Douyin, and WeChat Official Accounts. It emphasizes a clear process boundary of "titles and tags first, cover later," avoiding the generation of useless covers before titles are finalized. It does not handle automatic publishing, CRM integration, or video generation—it focuses solely on the production pipeline from article to titles, tags, cover, and publishing material packages, and can be migrated to other AI agents or GitHub repositories for further iteration.

Related Skills