eu-ai-act-fria
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user data regarding AI systems and operational contexts to generate legal reports, creating a vulnerability surface.
- Ingestion points: User input is requested throughout SKILL.md, particularly in 'Step 4', 'Step 5', and the 'Quick Question Set'.
- Boundary markers: The skill lacks explicit delimiters or instructions for the agent to distinguish between user-provided data and the skill's own operational guidelines.
- Capability inventory: The agent typically operates with tools for file management and shell execution, which could be exploited if the agent follows malicious instructions embedded in the input data.
- Sanitization: No input validation or sanitization mechanisms are defined for the user-provided content before it is interpolated into the workflow.
Audit Metadata