dogfood
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to navigate and interact with arbitrary external websites provided by the user, creating a surface for indirect prompt injection attacks.
- Ingestion points: Untrusted data enters the agent's context through
browser_snapshot,browser_vision, andbrowser_consolecalls within the 'Explore' and 'Collect Evidence' phases ofSKILL.md. - Boundary markers: There are no explicit instructions or delimiters used when the agent processes content from the browser to distinguish between page content and operational instructions.
- Capability inventory: The skill utilizes a suite of browser interaction tools (
browser_click,browser_type,browser_navigate) and performs file writing operations to the local filesystem to save screenshots and the final report. - Sanitization: No sanitization, filtering, or validation of the content retrieved from the web pages is performed before it is processed by the agent.
Audit Metadata