The Agent Skills Directory

[SAFE]: No security issues or malicious patterns were identified. The skill performs its stated function of extracting and validating URLs without accessing sensitive files, hardcoding credentials, or executing unauthorized code.- [PROMPT_INJECTION]: The skill processes untrusted user input and visits external links, creating a potential surface for indirect prompt injection. However, the instructions include explicit boundary markers that restrict the agent to extraction and validation tasks and prohibit detailed analysis of the content at the URLs, mitigating the risk of the agent being misled by malicious content found online. * Ingestion points: User-provided text in SKILL.md (Workflow Step 1). * Boundary markers: Present; instructions specifically limit the scope to extraction and validation and forbid content analysis. * Capability inventory: Uses the web_fetch tool to access external URLs. * Sanitization: None documented.

url-extractor