The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for utilizing the 'crawlberg' CLI tool. This tool is used to perform web crawling, capturing content, and managing request parameters like concurrency and rate limits. As the tool is aligned with the skill's stated purpose and authored by the same vendor, this is considered standard functionality.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from arbitrary external websites, creating a surface for indirect prompt injection attacks.
Ingestion points: External website content is captured via 'crawlberg crawl' as described in SKILL.md.
Boundary markers: The tool uses URL-based headers (e.g., '--- URL: ... ---') to delimit different pages in the output, which provides some structure but is not a robust defense against adversarial instructions within the page content.
Capability inventory: The agent has the capability to execute the 'crawlberg' command, read its output, and potentially perform further actions based on the crawled text.
Sanitization: There is no mention of sanitization, filtering, or safety checks applied to the captured web content before it is processed by the AI agent.

crawling-a-site