The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the requests Python library to perform network operations. This is a standard dependency for fetching web content.
[COMMAND_EXECUTION]: The skill executes a Python script (scripts/scrape_blogs.py) to discover feeds, parse XML content, and interact with the Apify API. This behavior is consistent with the skill's primary function.
[DATA_EXPOSURE]: The script interacts with the Apify API (api.apify.com) to perform scraping tasks. It properly handles the Apify API token by allowing it to be passed as a command-line argument or retrieved from the APIFY_API_TOKEN environment variable, which is a standard practice for managing credentials.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests untrusted data from external RSS feeds and blog URLs provided by the user.
Ingestion points: External XML feeds and HTML content are fetched via requests.get in scripts/scrape_blogs.py.
Boundary markers: The skill does not use specific delimiters or instructions to warn the agent about potentially malicious content embedded in the scraped blog posts.
Capability inventory: The skill can perform network operations via requests and output data to stdout, which may be processed by subsequent agent steps.
Sanitization: The skill performs basic XML parsing and client-side filtering but does not explicitly sanitize the text content for instructions directed at the LLM.
Note: This represents an inherent risk for any web-scraping tool rather than a specific malicious implementation.

blog-scraper