e-commerce-pflichten
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves a purely instructional purpose, providing legal frameworks (§ 312i, 312j, 312g, 355 BGB) for the agent to use when analyzing online store configurations. All external references point to official German legal repositories (gesetze-im-internet.de) or established legal databases (dejure.org).
- [INDIRECT_PROMPT_INJECTION]: The skill provides a template for processing user-supplied data regarding store checkout processes.
- Ingestion points: User-provided inputs describing shop types, checkout steps, and legal texts in SKILL.md and fact patterns in test.md.
- Boundary markers: Not explicitly defined in the prompt instructions.
- Capability inventory: No file system writes, network operations (beyond external links), or subprocess executions are defined within the skill.
- Sanitization: No specific input sanitization is implemented, but the lack of executable capabilities makes this surface low risk.
Audit Metadata