gws-modelarmor
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user-generated content and model responses for safety filtering.
- Ingestion points: Untrusted data enters the agent context via user prompts and model responses processed by the sanitization methods described in
SKILL.md. - Boundary markers: No specific delimitation or 'ignore embedded instructions' warnings are provided within the instruction file to isolate processed data from the agent's logic.
- Capability inventory: The skill invokes the
gwscommand-line tool to perform resource-specific methods (SKILL.md). - Sanitization: While the primary purpose of the skill is sanitization, the processing of untrusted input creates a surface for potential indirect prompt injection attacks.
Audit Metadata