The Agent Skills Directory

[SAFE]: The skill utilizes internal platform modules (contracts.web, kernel.web) to perform crawling tasks via a structured router (WebIntelligenceRouter). This indicates the skill is designed to operate within the agent's native environment constraints.
[SAFE]: The script correctly handles domain-based filtering through allowed_domains and blocked_domains parameters, which helps prevent unauthorized access to restricted network locations.
[SAFE]: No signs of obfuscation, hardcoded credentials, or persistence mechanisms were found in either the YAML configuration or the Python execution logic.
[SAFE]: The skill maintains a clear separation between user input (URLs) and the execution logic, reducing the risk of traditional command injection.

web-crawl