linking-embedding-framing
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to verify legal norms and case law by referencing official German and European legal portals, such as
gesetze-im-internet.de,dejure.org, andeur-lex.europa.eu. These are well-known and trusted official sources for legal research. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external evidence such as platform posts, contracts, and documents, creating a potential surface for indirect prompt injection.
- Ingestion points: The agent is directed to analyze 'Plattformpost', 'Vertragsurkunden', and 'Originaldateien' provided by the user or external sources in the 'Sofortstart' and 'Pr%flogik' sections.
- Boundary markers: No specific delimiters or boundary instructions are provided to separate external content from the prompt's instructions.
- Capability inventory: The skill focus is purely informational and legal-analytical; it lacks capabilities for code execution, file system modification, or network exfiltration.
- Sanitization: The instructions do not include steps for sanitizing or filtering the ingested data.
Audit Metadata