Back to skill
v1.0.0

hk-insurance-news

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 7:15 AM.

Analysis

This instruction-only skill coherently gathers public Hong Kong insurance news and drafts a WeChat-style weekly article, with minor review points around scheduled use and untrusted web sources.

GuidanceThis appears safe for drafting a weekly news article from public sources. Before using it, decide whether it should run manually or on a schedule, review all cited sources and images, and do not grant separate WeChat publishing credentials unless you add explicit approval controls.

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.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Rogue Agents
SeverityInfoConfidenceMediumStatusNote
SKILL.md
每周一早上10点,自动搜索、整理过去一周香港保险行业的重要动态

The skill describes periodic autonomous collection. This is disclosed and aligned with the weekly-report purpose, but users should notice if they configure recurring invocation.

User impactIf scheduled externally, the agent may perform regular web searches and draft reports without a fresh manual prompt each time.
RecommendationUse manual invocation or an explicit scheduler you control, and require review before any article is published.
Agent Goal Hijack
SeverityLowConfidenceMediumStatusNote
SKILL.md
主要香港本地媒体(如信报、经济日报)...(可选)专业保险新闻网站或论坛的讨论热点

The skill relies on third-party web and forum content. That is normal for news aggregation, but such content should be treated as source material rather than as instructions to the agent.

User impactMisleading webpages, forum posts, or prompt-injection style text could affect the generated summary if the agent over-trusts retrieved content.
RecommendationHave the agent cite sources, cross-check important claims, and ignore instructions embedded inside retrieved webpages or forum posts.