The Agent Skills Directory

[SAFE]: The skill follows its stated purpose of ingesting WeChat articles into a knowledge base using legitimate tools like BeautifulSoup4, MarkItDown, and PyMuPDF.
[COMMAND_EXECUTION]: The script fetch_article.py uses subprocess.run to execute curl for fetching article content. This is implemented securely using an argument list rather than a raw shell string, preventing shell injection. The operation is limited to fetching data from a user-provided URL and does not involve executing remote code.
[DATA_EXPOSURE_AND_EXFILTRATION]: No evidence of unauthorized data access or exfiltration. The script performs network operations solely to fetch the content requested by the user. Standard file writing is used to save the resulting Markdown files to a user-specified output directory.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external URLs and PDFs. While this presents an inherent attack surface for indirect injection if the processed content contains malicious instructions, the risk is categorized as low as the skill merely reformats the content into Markdown and does not provide an execution environment for the ingested data. Basic sanitization is applied to filenames and text cleanup.

wechat-article-ingest