code-documentation-code-explain
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code provided by the user to generate educational content, which creates a potential surface for instructions embedded in code comments or metadata to influence the agent's narrative.
- Ingestion points: User-supplied code and analysis requirements are ingested via the
$ARGUMENTSparameter inSKILL.mdand processed by the agent using logic described inresources/implementation-playbook.md. - Boundary markers: The instructions do not define explicit delimiters or 'ignore' directives to isolate the input code from the agent's control instructions.
- Capability inventory: The skill is primarily focused on generating text-based explanations and Mermaid diagrams; it does not request or use capabilities for network exfiltration, filesystem modification, or shell execution.
- Sanitization: No input sanitization or filtering logic is specified for the code provided to the agent for analysis.
Audit Metadata