The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill connects to an external API at https://onetotenvip.com/skill/cozeSkill/getWxDataByCategoryAndTime to retrieve article data. This domain is not an established or well-known service.
[EXTERNAL_DOWNLOADS]: The script scripts/fetch_articles.py implements a custom network fetcher (fetch_without_sni) that explicitly disables SSL certificate verification using ssl.CERT_NONE and sets check_hostname = False. This makes the connection highly susceptible to Man-In-The-Middle (MITM) attacks, allowing an attacker to intercept or modify the fetched article data.
[REMOTE_CODE_EXECUTION]: The script scripts/generate_hot_html.py performs unsafe dynamic content generation. It takes article titles, author names, and URLs from the external API and interpolates them directly into an HTML template using f-strings without any sanitization or escaping. This creates a Cross-Site Scripting (XSS) vulnerability if the API returns malicious payloads.
[COMMAND_EXECUTION]: The skill requires the execution of two local Python scripts (fetch_articles.py and generate_hot_html.py) to function. These scripts manage network requests and file system operations (writing HTML and JSON files).
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. It retrieves untrusted data from an external API and instructs the agent to display it "as-is" in a Markdown table. Malicious content within the article titles or metadata could be used to attempt to influence the agent's behavior or deceive the end-user.

wechat-original-article-king