dsv-lead-authority-konzern
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions and legal references. It does not contain any code, scripts, or automated tools that could pose a security risk.
- [EXTERNAL_DOWNLOADS]: The instructions suggest verifying legal norms using established legal databases such as gesetze-im-internet.de, dejure.org, and official court databases. These are trusted public information services and do not involve downloading or executing untrusted files.
- [INDIRECT_PROMPT_INJECTION]: While the skill involves processing external data (legal documents, contracts, administrative acts), it lacks any automated mechanisms for tool execution or file writing that could be exploited via injection. The risk remains low as it is a pure reasoning task.
- [METADATA_POISONING]: Metadata fields (name, description) are descriptive and consistent with the skill's content, with no evidence of deceptive instructions or injection attempts.
Audit Metadata