Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Dream Novel Writer
v1.0.1基于雪花写作法和三层记忆架构的小说创作辅助技能，支持从核心种子到章节产出的全流程小说架构与长文本记忆维持。包含了独立提示词库与 JS 管理脚本自动化工具。
⭐ 0· 44·0 current·0 all-time
byHalcyon@waitli
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The SKILL.md describes JS management scripts (./scripts/*.js) and resource templates (./resources/*.md) as integral parts of the feature set. The published package contains only SKILL.md and no scripts/resources. Additionally, the instructions assume Node.js is available (calls to node ./scripts/...), but required binaries list is empty. This is a clear inconsistency between the claimed purpose and the actual artifact.
Instruction Scope
Runtime instructions direct the agent to read from and write to local files (./resources/*.md, temp_outline.txt, chapter_X.md, JSON update files), run node scripts to assemble prompts and update state, and to silently iterate through entire-book generation without pausing for user confirmation. Those file operations and long-running silent behavior go beyond simple prompt guidance and require local scripts/resources which are absent from the bundle.
Install Mechanism
There is no install spec (instruction-only), which minimizes supply-chain risk. However, because the instructions require scripts that are not provided, the skill will either fail at runtime or rely on externally supplied files. The absence of an install step means nothing will be placed on disk automatically—so operational success depends on the runtime environment having the expected files and Node.js.
Credentials
The skill requests no environment variables or credentials, which is appropriate for a local authoring helper. That said, it implicitly requires Node.js and filesystem write access and will create/modify files in the project workspace; those requirements are not declared in metadata and should be made explicit.
Persistence & Privilege
The skill is not force-installed (always:false) and is user-invocable, which is normal. Still, the instructions demand autonomous, uninterrupted generation and persistent storage of project data (creating workspaces, repeatedly updating JSON and markdown files). If the agent is allowed to invoke skills autonomously, this could lead to extensive automated file writes without repeated user consent—consider limiting or requiring per-chapter confirmation.
What to consider before installing
This skill's README says it relies on ./scripts/*.js and ./resources/*.md and runs node to manage state, but the published package contains only SKILL.md. Before installing or using it: (1) don't assume the scripts exist—verify the origin and obtain the actual scripts/templates from a trusted source; (2) ensure Node.js is available and inspect any scripts you will run; (3) be aware the skill instructs the agent to silently generate and write an entire novel (many files) without per‑chapter confirmation—if you want control, require the agent to pause for your approval between chapters; (4) if you can't verify or review the missing scripts/resources, treat the skill as untrusted and avoid giving it filesystem write access. Absence of automated scan hits does not mean the skill is safe—these mismatches are the main concern.Like a lobster shell, security has layers — review code before you run it.
latestvk97ehvtc9t5bv7qt15eje1b8gn842j0m
License
MIT-0
Free to use, modify, and redistribute. No attribution required.