ClawHub

Ai News Collectors 1.0.0

v1.0.0

AI 新闻聚合与热度排序工具。当用户询问 AI 领域最新动态时触发，如："今天有什么 AI 新闻？""总结一下这周的 AI 动态""最近有什么火的 AI 产品？""AI 圈最近在讨论什么？"。覆盖：新产品发布、研究论文、行业动态、融资新闻、开源项目更新、社区病毒传播现象、AI 工具/Agent 热门项目。输出中文...

⭐ 1· 223·2 current·2 all-time
bykuang k@kkblock
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Skill name/description (AI news aggregation and ranking) match the SKILL.md and the provided sources list. There are no unexpected binaries, env vars, or install steps required that would be unrelated to collecting news.
ℹ
Instruction Scope
Instructions are explicit about running 8–12 layered web searches across six dimensions and fetching full texts (it mentions using a web_fetch step). This stays within news-aggregation scope. Note: web_fetch implies the agent will retrieve external URLs and full article contents (including paywalled pages which the skill says to mark as 'needs subscription'), so expect significant external network activity. The instructions do not request reading local files or environment secrets.
✓
Install Mechanism
No install spec and no code files — instruction-only skill. This minimizes disk/execution risk because nothing is downloaded or installed.
✓
Credentials
Skill requires no environment variables, credentials, or config paths. All declared requirements are proportional to an aggregator: public web access and known news/community sources.
✓
Persistence & Privilege
always:false and normal autonomous invocation settings. The skill does not request persistent system-wide privileges or attempt to modify other skills or configs.
Assessment
This skill looks coherent and does what it says: run many targeted web searches and fetch articles from the listed sources. Before installing, decide whether you are comfortable allowing the agent to browse and fetch full web pages (the SKILL.md expects web_fetch). If you want tighter control: (1) restrict the agent's browsing capability or limit it to a curated allowlist of sources in references/sources.md; (2) disallow fetching paywalled content or block domains you don't trust; (3) consider keeping human approval for publishing the aggregated output. There are no obvious credential or install risks, but expect higher external network activity and potential exposure to low-quality community posts — review the sources list if you need stricter source trust controls.

Like a lobster shell, security has layers — review code before you run it.

latestvk977fban542pwkzrvaadap0yg982xzbj

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments