gpsr-sicherheitspruefung
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, user-provided data which introduces a potential surface for indirect prompt injection.
- Ingestion points: The skill accepts several user-supplied inputs in
SKILL.md, including product designations, functions, and target user descriptions. - Boundary markers: The prompt instructions do not include specific delimiters or warnings to the agent to ignore instructions that might be embedded within these user inputs.
- Capability inventory: The skill utilizes a multi-agent architecture (researcher, drafter, reviewer) with no restrictions on tool usage in the YAML frontmatter, implying access to standard shell and file system capabilities.
- Sanitization: There is no evidence of input validation or sanitization routines to filter out malicious payloads within the provided product information.
Audit Metadata