Agent Skills: story-control-panel

Define what the Story Control Panel shows, how to compute signals, and how the author should act on alerts.

UncategorizedID: aevatarAI/aevatar-agent-framework/story-control-panel

Install this agent skill to your local

pnpm dlx add-skill https://github.com/aevatarAI/aevatar-agent-framework/tree/HEAD/experimental/novel/agent_skills/story-control-panel

Skill Files

Browse the full folder contents for story-control-panel.

Download Skill

Loading file tree…

experimental/novel/agent_skills/story-control-panel/SKILL.md

Skill Metadata

Name
story-control-panel
Description
Define what the Story Control Panel shows, how to compute signals, and how the author should act on alerts.

When to use

  • Author asks: “我现在这本书/这个 story 的状态怎样?后面哪里危险?哪些伏笔欠债?设定是否失控?”

Inputs (SSOT)

  • Chapter text: .txt
  • Other artifacts: .md(大纲/台账/测试报告/偏离影响分析/会话日志)
  • SQLite: index only (rebuildable)

Output

  • A dashboard-ready summary (for UI) with:
    • Progress (Volume/Story/Chapter)
    • Alerts (timeline/canon/test failures)
    • Hotspots (canon rule usage heat, repeated confusion points)
    • Backlog (setup/payoff debts)

Signals (v1 minimal)

  • Progress
    • Chapters drafted / reviewed / finalized
  • Hard Alerts (must block auto-continue)
    • Timeline contradiction
    • Canon rule violation (explicitly tested)
    • Narrative Tests failed (severity=ERROR)
  • Soft Alerts (should suggest options)
    • Too many unresolved setups (debt ratio)
    • Persona “confusion” score spikes

Procedure (how to reason)

  1. Start from hard alerts → locate where it breaks → propose minimal fix vs branch rewrite.
  2. Then handle debt/backlog → propose payoff roadmap.
  3. Finally handle optimization signals (persona/style/pacing).