c4-architecture-c4-architecture

Pass

Audited by Gen Agent Trust Hub on Aug 16, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The workflow is designed to discover and analyze sensitive infrastructure and deployment definitions, including Dockerfiles, Kubernetes manifests, and Terraform configurations. While the skill primarily saves this information to a local directory, the automated aggregation of these sensitive details represents a data exposure risk if the resulting documentation is subsequently shared, committed to a public repository, or accessed by unauthorized agents.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the target codebase and passes it to subagents for analysis and synthesis without implementing boundary markers. This creates a surface for indirect prompt injection where malicious instructions hidden in the analyzed code could potentially manipulate the output of the documentation process.
  • Ingestion points: Every file and subdirectory within the target repository is analyzed bottom-up by specialized agents (SKILL.md, Sections 1.2 and 4.2).
  • Boundary markers: There are no explicit delimiters or instructions provided to the subagents to ignore embedded prompts within the source material.
  • Capability inventory: The skill generates and writes numerous documentation files to the local filesystem and creates OpenAPI/Swagger specifications.
  • Sanitization: No sanitization or filtering of the ingested source code is performed before it is analyzed by the agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 16, 2026, 01:29 AM
Security Audit — agent-trust-hub — c4-architecture-c4-architecture