story-pipeline

What This Skill Does

The Story Pipeline skill provides a robust, professional-grade framework for long-form narrative generation within OpenClaw. Designed for authors, content creators, and world-builders, it moves beyond simple chat-based writing by maintaining a persistent 'story state.' By utilizing an internal graph management system, the skill tracks character traits, scene histories, and narrative hooks, ensuring that your story maintains consistency over dozens or even hundreds of episodes. The core of this skill is a dual-layered verification mechanism: first, an automated AI quality check evaluates plot coherence, pacing, and emotional impact against a scoring threshold (requiring at least 7/10 to proceed). Second, the skill pauses for human intervention, providing a clear preview and control options, allowing you to approve, modify, or pivot the narrative direction before the next segment is written.

Installation

To integrate this into your agent, run the following command in your terminal: clawhub install openclaw/skills/skills/hexidyg/shorts-builder Ensure your OpenClaw environment has adequate write permissions to the /data directory, as the skill periodically saves progress to pipeline_state.json.

Use Cases

• Web Novel Series: Generate multi-chapter serials with complex subplots that remain internally consistent.
• Roleplay Management: Maintain evolving character relationships and background lore for table-top or AI-driven roleplay sessions.
• Screenwriting: Draft scene-by-scene blueprints that require strict narrative pacing and hook management.
• Educational Storytelling: Develop interactive lessons or historical narratives that follow a predictable, structured progression.

Example Prompts

1. "Start a new story pipeline, theme: Cyberpunk detective noir in Neo-Tokyo, target episodes: 50."
2. "Continue story pipeline [pipeline_id]"
3. "Modify: Make the protagonist appear more skeptical of the informant's claims in this episode."

Tips & Limitations

• Consistency: Always check the graph status if the AI forgets a minor detail; you can use the modify feature to 'inject' lost context back into the current episode.
• Pacing: If you find the story moves too fast, adjust your initial prompt to request 'slower pacing' or 'longer scene descriptions'.
• Data Privacy: Be aware that the pipeline_state.json stores your full story content locally. Keep backups if you are working on highly sensitive creative intellectual property.
• Persistence: Always remember to 'Pause' if you intend to take a long break to ensure state is cleanly serialized.