teach
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill logic is confined to providing educational explanations and does not perform any dangerous actions like unauthorized file access or network communication.
- [INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface for untrusted data as it processes user conversation and source code to generate its explanations.
- Ingestion points: Processes user conversation and existing code files as described in
SKILL.md. - Boundary markers: No specific boundary markers or instructions to ignore embedded commands are defined for the processed data.
- Capability inventory: The skill is restricted to generating text and diagrams; it does not possess capabilities for file modification, network exfiltration, or shell execution.
- Sanitization: No input sanitization or validation logic is present.
Audit Metadata