colorize
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it processes external files or code specified via user arguments.
- Ingestion points: External UI code files passed through
$ARGUMENTSinSKILL.md. - Boundary markers: Absent; there are no explicit boundary delimiters or formatting constraints to isolate the ingested file content from the agent instructions.
- Capability inventory: The skill possesses file modification capabilities ("edit code directly").
- Sanitization: Absent; no validation or escaping mechanisms are specified for the untrusted UI code input.
Audit Metadata