novel-multi-agent-skill
PassAudited by ClawScan on May 10, 2026.
Overview
This appears to be a user-directed local novel-writing workflow with no evidence of hidden data theft or destructive behavior, but it uses shell scripts and relies on separately configured agents.
Before installing, review the bash scripts, run them only in a dedicated novel project directory, and remember that full automation depends on agent configurations you create separately. Avoid putting sensitive personal information into project files if those files may later be sent to a model provider through OpenClaw.
Findings (2)
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.
Running the workflow can create or overwrite files in the selected project folder.
The script creates directories and writes project files to a path supplied by the user. This is expected for a project generator, but it can modify local files if the user points it at an existing or sensitive directory.
mkdir -p "$PROJECT_DIR" ... "$TEMPLATES_DIR/00-总纲模板.md" > "$PROJECT_DIR/00-总纲.md"
Run it only from a trusted copy of the skill and use a new, dedicated project directory.
The reviewed package does not fully determine what the external/user-configured agents will do.
The complete automated behavior depends on separate agent configurations created by the user, so those agent prompts and settings are outside the reviewed artifacts.
要使用完整的自动化功能，需要配置以下5个智能体... 在 OpenClaw 中创建对应的代理，配置相应的系统提示词。
Review or author the five agent configurations yourself and avoid importing untrusted agent prompts.