novel-writing
PassAudited by ClawScan on May 1, 2026.
Overview
This appears to be a local novel-writing helper that creates and updates project files, with no evidence of network access, credential use, or hidden behavior.
This skill looks safe for its stated purpose. Before using it, understand that it creates and updates local novel project files and stores feedback/context for later writing; keep backups and do not paste private or untrusted feedback unless you are comfortable with it influencing future drafts.
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.
If invoked with the wrong chapter or content, the tool could overwrite or change local novel project files.
The CLI can write chapter content and update project state files. This is aligned with the novel-writing purpose, but it is still local file mutation.
chapter_file.write_text(content, encoding="utf-8") ... state_file.write_text(json.dumps(state, ensure_ascii=False, indent=2), encoding="utf-8")
Run commands only in the intended workspace and keep backups of important chapters or outlines.
Stored feedback could influence future generated writing if it is treated too strongly by the agent.
The skill intentionally stores reader feedback and reloads it during later writing. This is useful for the stated purpose, but reader feedback may contain untrusted or misleading text.
反馈入库 `novel feedback <章节> <反馈内容>` ... 将读者反馈存入追踪器，写相关章节时自动加载。
Review stored feedback periodically and treat it as reference material, not as instructions that override the author’s intent.
Users have less external context for verifying the publisher or project history before use.
The package has limited provenance information and no installation instructions, though the included behavior is local and purpose-aligned.
Source: unknown; Homepage: none; No install spec — this is an instruction-only skill.
Inspect the included files before use and install or run only from the trusted ClawHub package context.