Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed Chinese news aggregation skill that fetches fixed RSS and webpage sources, with content-integrity and untrusted-content risks users should understand before relying on it.
Install only if you want a Chinese-market news aggregation helper that contacts third-party RSS/news sites. Treat fetched article text and links as untrusted, verify market-moving claims from authoritative sources, and prefer an updated version that replaces HTTP feeds with HTTPS and labels remote content clearly.
scripts/news_market.py:112Untrusted remote news content is passed into AI-facing output without a trust boundary
scripts/news_market.py:39News feeds are retrieved over unauthenticated plaintext HTTP
The skill is described as focused on China A-share market news, but the documented behavior spans broad politics, technology, general media, and government sources. This mismatch is dangerous because it can bypass user or platform expectations, causing the skill to access and surface content outside its stated scope, which undermines trust and can evade policy review based on the narrower description.
The skill performs external network access to RSSHub mirrors and potentially additional web fetching, but the manifest does not explicitly declare tool scope or permissions. This creates a transparency and governance gap: users and hosting platforms may not realize the skill can reach arbitrary external content, which increases the risk of unexpected data egress, policy violations, or unsafe remote content retrieval.
The description states the skill is focused on the China A-share market, and later guidance says English content may be translated, but the document does not offer the user any language or locale choice. Under SQP-3, locale/language constraints should be opt-in or clearly justified; here the skill appears to impose a China/Chinese-oriented scope by default.
The module docstring labels the tool as a Chinese news aggregator, and the command descriptions, help text, and runtime messages are all presented only in Chinese. Under the stated policy, forcing a specific language without opt-in or an explicit documented locale justification is a natural-language policy violation.
The manifest says the skill is dedicated to querying Chinese A-share market information, yet the configured sources include broad outlets such as 人民网时政, 36氪, 观察者网/澎湃, IT之家, InfoQ, and general government policy feeds. This makes the implemented behavior a general Chinese news aggregator with some finance sources, which materially exceeds the stated A-share-market focus.
The markdown instructs the agent to retrieve full article content via RSS or web scraping, but it does not clearly warn that this requires live network access to third-party sites and may ingest untrusted remote content. That omission increases the chance of silent external requests, privacy surprises, and unsafe handling of fetched content in downstream agent workflows.
This code file performs live network requests to external domains using urllib, including direct fetches to numerous RSS and web endpoints. Although stderr progress messages are present, they do not disclose the privacy/safety implication that running the tool contacts third-party services and transmits request metadata such as IP address and user-agent.
The help epilog advertises category --cat tech and category --cat finance, but the parser only allows categories from SOURCES.keys(), namely hot, tech_media, securities, and official. This is an active contradiction between the inline documentation and executable behavior.
No suspicious patterns detected.