The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the agent's bash capabilities to execute commands such as curl, cat, wc, and head for fetching and inspecting file contents. It includes explicit instructions in SKILL.md to normalize user-provided domains before use to mitigate command injection risks.
[EXTERNAL_DOWNLOADS]: Uses curl to make outbound HTTP requests to user-specified domains to retrieve robots.txt, llms.txt, and llms-full.txt. These operations are transparently disclosed in the documentation and are essential to the skill's primary purpose.
[PROMPT_INJECTION]: Indirect Prompt Injection Surface: The skill fetches and reads content from external domains. If these files contain malicious instructions, they could potentially influence the agent's behavior during the audit process. However, this is an inherent risk of any tool designed to process web content, and no specific malicious patterns were found in the skill's logic.

llms-txt-checker