eu-ai-act-compliance
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
rg(ripgrep) command-line utility to search for specific keyword patterns within a user-provided codebase. This is used to discover AI models, SDKs, and logic related to biometrics or high-risk decision-making. This is a functional requirement for the skill's purpose as a code auditor. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of code repositories and technical documents provided by the user for review.
- Ingestion points: File contents from code repositories, model cards, technical documentation, and AI policies are read into the agent's context as evidence for the compliance review.
- Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions within the artifacts being reviewed.
- Capability inventory: The skill utilizes shell-based search capabilities (
rg) and has the authority to write report files (.md) to the local filesystem. - Sanitization: The skill includes a clear safety directive in
SKILL.mdto redact real personal or confidential values, replacing them with generic categories in the final report. This reduces the risk of data exposure.
Audit Metadata