Back to skill
Skillv1.0.0
ClawScan security
Writing Style Cloner - 个人写作风格克隆器 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 9, 2026, 2:58 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This is an instruction-only skill whose declared purpose (rewriting transcripts/drafts into Antonia's personal social-media style) matches the files and runtime instructions; it does not request credentials, installs, or unusual privileges.
- Guidance
- This skill is internally coherent and low technical risk because it's purely a rules/prompts repository. Before using it, consider: (1) privacy — transcripts may contain sensitive or personal data, so scrub or review inputs/outputs before publishing; (2) consent/ethics — cloning a named person's distinctive style can raise impersonation or attribution concerns, obtain permission or avoid presenting output as literally authored by that person; (3) copyright/plagiarism — the SKILL.md states it’s based on historical articles; avoid verbatim reuse of copyrighted passages from originals; (4) quality/safety — test outputs on representative samples and verify they preserve facts correctly and don’t introduce hallucinated claims. If you need the agent to call external services or access files later, re-evaluate for required credentials or broader privileges.
Review Dimensions
- Purpose & Capability
- okName/description (style-cloning for 'Antonia') align with the included SKILL.md, README, and reference docs. The skill is instruction-only and does not declare any unrelated environment variables, binaries, or config paths—nothing requested appears disproportionate to the stated purpose.
- Instruction Scope
- okSKILL.md contains prescriptive rewrite rules, style guidelines, and an explicit output format; it instructs the agent only how to transform user-provided text. It does not direct the agent to read system files, call external endpoints, or access environment variables beyond user input. Note: the rules require preserving core facts from the user's transcript — that means transcripts may include PII or sensitive content, so outputs should be reviewed before publishing.
- Install Mechanism
- okNo install spec and no code files that would be written to disk; lowest-risk delivery model (instruction-only).
- Credentials
- okNo required environment variables, credentials, or config paths are declared. The README suggests optional parameters (word count, platform) passed as user input, which is appropriate and proportional.
- Persistence & Privilege
- okalways is false and the skill is user-invocable; it does not request persistent system presence nor modify other skills. Autonomous invocation is allowed by default but this skill's scope is limited to text transformation.