c4-architecture-c4-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted codebase files which creates a surface for indirect prompt injection attacks.
- Ingestion points: The skill identifies and analyzes all subdirectories and files in a repository, including source code, READMEs, and deployment configurations (Sections 1.1, 1.2, 3.1, 4.1).
- Boundary markers: The prompts for sub-agents (c4-code, c4-component, etc.) use block-text formatting to interpolate directory paths and file lists, but they do not include explicit instructions for the agent to ignore or treat text within the analyzed files as data rather than instructions.
- Capability inventory: The agent uses codebase search, file read operations, and file write operations to create documentation in the
C4-Documentation/directory. - Sanitization: There is no mention of sanitization, filtering, or escaping of the code content before it is processed by the documentation sub-agents.
Audit Metadata