The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches titles, URLs, and metadata from external websites (sitemaps, RSS feeds, and HTML pages) that are not under the user's control. If these resources contain malicious text, they could influence the agent's behavior during subsequent analysis of the catalog.
Ingestion points: The catalog_domain function in scripts/catalog_content.py retrieves data from sitemap.xml, RSS feeds, and blog index pages using the requests library.
Boundary markers: The skill does not implement delimiters or safety instructions to distinguish external content from agent instructions.
Capability inventory: The script has network access and can write to the local file system.
Sanitization: No sanitization is applied to the extracted web content before it is provided to the agent.
[REMOTE_CODE_EXECUTION]: The skill integrates with the Apify platform using a specific third-party actor (onescales~sitemap-url-extractor) for sitemap extraction. This involves executing unverified logic on a remote service via an API token.
[COMMAND_EXECUTION]: The CLI tool accepts an --output parameter that specifies a local file path for saving results. The script does not validate this path, potentially allowing the agent to overwrite sensitive files if a malicious path is provided.

site-content-catalog