skills/klotzkette/claude-fuer-deutsches-recht/buchhaltung-konten-kanzlei-erechnung/Gen Agent Trust Hub
buchhaltung-konten-kanzlei-erechnung
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources such as invoices, bank statements (CSV, CAMT, MT940), and payment advisories. This creates a surface for indirect prompt injection where an attacker could embed malicious instructions in fields like the payment purpose or invoice descriptions.
- Ingestion points: Processes external files including CSV, CAMT, MT940, PDF, and payment advisories as listed in the 'Datenquellen' section.
- Boundary markers: No specific delimiters or instructions to ignore embedded commands are defined for processing external data.
- Capability inventory: The skill performs payment matching logic, calculates aging for debits, and generates output for payment reminders and matching registers.
- Sanitization: The instructions do not specify any sanitization, escaping, or validation techniques for the content extracted from external financial documents.
- [SAFE]: The skill includes positive security practices in its 'Leitplanken' (Guardrails) section, explicitly instructing the agent not to store sensitive credentials such as Bank PINs, TANs, or API secrets, and requiring separate authorization for payment orders.
Audit Metadata