vehicles-at-speed

Pass

Audited by Gen Agent Trust Hub on Nov 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs the agent to download and analyze media files (MP4, JPG) and social media content from external sources such as goodcase.ai and x.com. These resources are not from trusted repositories or organizations.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by requiring the agent to use vision or browser tools to inspect external media and web pages. This creates a potential vector for instructions embedded in the external content to influence the agent's behavior.
  • Ingestion points: Links in references/cases.md to external media and original social media posts.
  • Boundary markers: No boundary markers or 'ignore' instructions are provided to help the agent distinguish between reference content and its own operating instructions.
  • Capability inventory: The skill utilizes image, video, and browser tools to inspect content and generates complex textual prompt artifacts.
  • Sanitization: There is no specified mechanism for sanitizing or validating the content retrieved from external URLs.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 15, 2026, 03:36 AM