Story Long Write
PassAudited by ClawScan on May 12, 2026.
Overview
This is a coherent instruction-only novel-writing skill, with the main risks being normal but noticeable project file creation, persistent story notes, and an optional helper-agent handoff.
This skill appears safe for its stated purpose. Install it if you want a file-based long-novel writing workflow, but run it inside a dedicated project folder, review the files it creates, and only enable the optional story-explorer helper if you trust that helper agent.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent may create or update many writing-project files rather than only replying in chat.
The skill directs the agent to write generated chapters directly into project files. This is expected for a file-based writing workflow, but it is still local file mutation that users should understand.
每章写完直接写入 `正文/` 目录，不要先输出到对话
Use a dedicated novel project folder, keep backups or version control, and review generated files before relying on them.
Mistakes or unwanted information in tracking files may be reused in future chapters or outlines.
The skill uses persistent project context files across writing sessions. This supports continuity but means stale, incorrect, or user-edited context can influence later outputs.
`追踪/上下文.md` | 全书 | Phase 4 首次日更（workflow-daily 自动创建） | 每次日更开始时
Periodically review the tracking files, especially 上下文.md, 伏笔.md, and 时间线.md, and remove inaccurate or sensitive content.
Another local agent may read or summarize the writing project and its output may influence the chapter being written.
The skill may hand the project directory and context-loading request to a separate story-explorer agent if present. That is coherent for project navigation, but the helper agent itself is not included in the reviewed artifacts.
可 spawn `Agent(subagent_type: "story-explorer", prompt: "项目目录：{dir}\n查询类型：context_load\n查询参数：准备写第 {N} 章")` 一次获取上下文Only use this optional helper if the story-explorer agent is trusted, or ask the assistant to load context directly without spawning a subagent.