od-prototype-mockup-device-3d
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input via the
$ARGUMENTSvariable and embeds it directly into a brief that determines the agent's file-authoring behavior. - Ingestion points: SKILL.md (via the
$ARGUMENTSvariable inside the brief description) - Boundary markers: Absent. The user input is interpolated directly into the string without delimiters or instructions to ignore embedded commands.
- Capability inventory: The skill explicitly directs the agent to "Author the files yourself with your own file-editing tools", giving the agent permission to write to the filesystem based on the instructions derived from the brief.
- Sanitization: Absent. There is no evidence of validation, escaping, or filtering of the content provided in
$ARGUMENTS.
Audit Metadata