kaltstart-routing-triage
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 external data such as incoming legal correspondence and uploaded files, which is a standard attack surface for indirect prompt injection.
- Ingestion points: The agent is instructed to read incoming correspondence, emails, and uploaded document contents to perform its triage tasks (SKILL.md, Section 1.1).
- Boundary markers: The instructions do not specify technical delimiters (like XML tags or clear separators) to isolate external input from the agent's core instructions.
- Capability inventory: The skill drafts legal documents, calculates statutory deadlines, and prepares billing statements based on processed data.
- Sanitization: While no specific input sanitization is mentioned, the skill mandates explicit human review ('Freigabe') for critical outputs, which acts as a manual control against prompt injection effects.
Audit Metadata