story-pipeline

What This Skill Does

The story-pipeline skill is a powerful framework designed for automated, serialized content creation within OpenClaw. It functions as a specialized workflow engine that maintains narrative continuity across multiple episodes. By leveraging an integrated graph management system, the skill tracks characters, locations, and narrative hooks, ensuring that every subsequent episode builds logically upon previous ones. It features a sophisticated dual-control mechanism: an AI-driven quality assurance layer that scores narrative flow and consistency, followed by a human-in-the-loop confirmation gate that ensures the generated output aligns perfectly with the user's creative vision.

Installation

To integrate this skill into your environment, run the following command in your terminal: clawhub install openclaw/skills/skills/hexidyg/shorts-builder-cn Ensure you have the required API endpoints configured for the graph management module as specified in the technical documentation.

Use Cases

This skill is ideal for content creators, screenwriters, and creative agencies. It excels at:

• Generating long-form serial scripts where character development and plot continuity are paramount.
• Managing complex narrative threads, such as "hook-heavy" web series or dramatic storytelling where subplots must be closed over multiple episodes.
• Automated prototyping of episodic content that requires consistent quality checks before finalization.

Example Prompts

1. "开始一个新的剧情管道，主题：赛博朋克下的侦探故事，目标集数：5集"
2. "继续剧情管道 pipeline_2026-03-05-001"
3. "修改：调整第三集的反转情节，让主角不要直接发现线索，而是通过侧面描述引导"

Tips & Limitations

• AI QA: The AI reviewer score is a soft constraint; if the quality falls below 7, the system automatically triggers a rewrite. Ensure your system prompts for the generator are descriptive to minimize unnecessary retries.
• Graph Persistence: The system relies on external API calls for graph management. If your network connection is unstable, cached states in data/pipeline_state.json are essential for resuming progress.
• Human Intervention: The pipeline will always pause at the confirmation gate. Do not forget to review the output, as the final narrative quality depends on your feedback during the manual confirmation phase. High-quality feedback loops will significantly improve the AI's future performance for that specific pipeline.