The Agent Skills Directory

[SAFE]: The skill defines procedures for automated QA testing, including smoke tests, interaction tests, visual regressions, and accessibility audits. All operations are consistent with the stated purpose of a browser interaction verification tool.- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes untrusted data from external websites. 1. Ingestion points: External target URLs, console logs, and DOM content from navigated pages. 2. Boundary markers: None explicitly defined in the skill instructions to delimit external content. 3. Capability inventory: URL navigation, element interaction (clicking links/buttons), and form submission. 4. Sanitization: No specific sanitization or filtering of external website content is defined before processing.

browser-qa