dsv-verdacht-vs-festgestellt
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured instructions for professional legal assessment of GDPR Art. 33 (data breach notification). No evidence of prompt injection, data exfiltration, or persistence mechanisms was found.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns were identified. The instructions mention verifying legal norms via official and well-known German legal databases (gesetze-im-internet.de, dejure.org, openJur), which is standard for legal research.
- [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or scripts that execute operations on the host system.
- [DATA_EXFILTRATION]: There are no hardcoded credentials or network operations that attempt to send data to external or untrusted servers.
- [INDIRECT_PROMPT_INJECTION]: The skill involves processing external data sources such as log entries, third-party reports, and legal documentation to assess breach status. While this is an ingestion surface, the skill does not possess any dangerous capabilities (like file writing or network requests) that could be weaponized by malicious content within those documents.
Audit Metadata