screenshots-marketing
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill file contains no executable scripts or automated commands. It serves as a discovery entry in the platform's catalog, providing metadata and links to an external GitHub repository.
- [INDIRECT_PROMPT_INJECTION]: Because the skill's purpose is to take screenshots of external websites using Playwright, it inherently creates a surface for indirect prompt injection. Malicious instructions could be embedded in the HTML or visual content of a targeted webpage to influence the agent's behavior during the screenshotting process.
- Ingestion points: External web content and URLs processed by the Playwright tool.
- Boundary markers: None defined in the current skill stub.
- Capability inventory: Automated browser interaction and screenshot generation via Playwright.
- Sanitization: No specific sanitization or filtering of the processed web data is mentioned.
Audit Metadata