architecture-guard
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill requires the agent to ingest content from
.context/architecture.md, which creates a vulnerability surface for indirect prompt injection. 1. Ingestion points: .context/architecture.md (Step 1). 2. Boundary markers: No delimiters or ignore instructions specified for the architecture file content. 3. Capability inventory: The skill triggers file system operations including creation and movement of files, and invokes secondary skills like context-sync. 4. Sanitization: No validation or escaping of the external file content is implemented.
Audit Metadata