The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration instructs the host to execute the @crawlbrulee/mcp package using the npx command. This process downloads and runs the official package from the NPM registry. Since this originates from a well-known service and belongs to the skill's authoring vendor, it is considered a legitimate dependency for the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of content from external websites via the scrape and map tools, which creates a potential surface for indirect prompt injection where malicious instructions on a webpage could influence the agent's behavior.
Ingestion points: The scrape, map, and scrape_result tools fetch and process content from user-provided or discovered URLs.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are defined in the skill's documentation, relying on the host environment's default handling.
Capability inventory: The tools enable web content retrieval, site mapping (URL discovery), and support for user-defined webhook callbacks for asynchronous jobs.
Sanitization: The tool provides options for data sanitization, including the cleanup parameter to remove specific HTML elements (like ads or popups) and the ability to convert output to Markdown, which helps mitigate risks associated with processing complex HTML.

crawlbrulee-mcp