lernmethoden

Pass

Audited by Gen Agent Trust Hub on Nov 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project context and materials provided by the user to design educational interventions, which represents a surface for indirect prompt injection. The mandatory evidence chain is as follows: 1) Ingestion points: Context is gathered in SKILL.md (Lernkontext erfassen) and referenced project materials. 2) Boundary markers: The skill encourages 'kognitive Pausen' (cognitive pauses) and human oversight to prevent thoughtless adoption of AI outputs. 3) Capability inventory: The skill is instructional and does not perform subprocess calls, file writes, or network operations during execution. 4) Sanitization: The skill explicitly recommends input sanitization and deterministic checks in references/runtime-ai.md.
  • [EXTERNAL_DOWNLOADS]: The documentation mentions installation via the Skills CLI from the author's GitHub repository. These references target the identified vendor 'siberius' and are consistent with standard tool distribution practices.
  • [SAFE]: The skill primarily consists of pedagogical instructions and security best practices. It explicitly advises against hardcoding credentials, such as API keys in client-side code, and encourages data minimization to protect user privacy.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 14, 2026, 08:48 PM