skills/klotzkette/claude-fuer-deutsches-recht/rechtsabteilung-open-source-copyleft-im-saas-stack/Gen Agent Trust Hub
rechtsabteilung-open-source-copyleft-im-saas-stack
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which creates a potential surface for indirect prompt injection.
- Ingestion points: The skill instructions in
SKILL.mddirect the agent to analyze external documents including contracts, emails, customer communication, and logs. - Boundary markers: There are no explicit instructions or delimiters defined to help the agent distinguish between its core instructions and potentially malicious content embedded in the processed documents.
- Capability inventory: The skill is restricted to text generation (memos, internal directives, and legal drafts) and does not utilize dangerous capabilities like shell execution or network exfiltration tools.
- Sanitization: No sanitization or filtering steps are defined for the ingested data.
- [SAFE]: The skill references reputable and official legal information services.
- The instructions include lookups on well-known German legal databases such as
gesetze-im-internet.de,dejure.org, andopenJurto verify legal norms. These are recognized as safe, authoritative sources for the intended legal use case.
Audit Metadata