skills/klotzkette/claude-fuer-deutsches-recht/mandantenakte-anlegen-mandantenbrief-vorlagen/Gen Agent Trust Hub
mandantenakte-anlegen-mandantenbrief-vorlagen
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted client information and case descriptions which are interpolated into persistent configuration files and legal documents.
- Ingestion points: User-provided client metadata, mandate scope, and evidentiary details are processed in SKILL.md.
- Boundary markers: The skill lacks explicit instructions or delimiters to distinguish untrusted client data from system instructions during file generation.
- Capability inventory: The agent is instructed to read and write to the filesystem, specifically creating directories and YAML/MD files within the ~/.claude/plugins/config/ directory.
- Sanitization: There are no defined mechanisms for validating or escaping input data to prevent characters from being interpreted as instructions or formatting controls in the resulting files.
Audit Metadata