od-prototype-social-youtube-thumbnail
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill directly interpolates user-provided arguments into a brief for the open-design MCP server tool, creating an attack surface for indirect prompt injection.
- Ingestion points: User input enters the system via the $ARGUMENTS variable in SKILL.md.
- Boundary markers: The skill lacks delimiters or explicit instructions to treat the $ARGUMENTS content as untrusted data.
- Capability inventory: The agent uses MCP tools to prepare designs and its own file-editing tools to author project files based on the brief's output.
- Sanitization: No sanitization or filtering is performed on the user-provided arguments.
Audit Metadata