ad-creative
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input that is then rendered within a browser environment, creating a vulnerability to script injection.
- Ingestion points: Headline text and body text gathered in Step 1 of
SKILL.md. - Boundary markers: None. The instructions do not define delimiters or protective instructions for the agent when processing these inputs.
- Capability inventory: The skill uses Playwright MCP to navigate to and render local files via
file://URLs. - Sanitization: None. There is no mention of escaping or validating HTML content before rendering.
- [DYNAMIC_EXECUTION]: The skill generates executable HTML content at runtime by interpolating user-provided strings into predefined templates and saving them as
ad-{slug}.html. - [COMMAND_EXECUTION]: The skill provides configuration guidance for Playwright MCP that includes the
--allow-unrestricted-file-accessflag. This configuration recommendation reduces the security isolation of the browser environment by allowing it to read any file on the local system.
Audit Metadata