architect
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests external source code (via the 'how' skill) and user tasks to generate architectural designs. However, this is inherent to the skill's primary function of code analysis and design.
- Ingestion points: Grounding artifacts in Phase A and user-provided task descriptions.
- Boundary markers: The skill does not explicitly define delimiters to isolate external content within its design prompts.
- Capability inventory: Spawns sub-agents via the 'arena' skill and writes design rationale and code sketches to the local environment.
- Sanitization: No specific input sanitization is documented.
- [COMMAND_EXECUTION]: The skill mentions a local maintenance script used for configuration management.
- Evidence: The 'Models' section provides instructions to 'rerun tools/generate.mjs' to update role defaults from internal configuration files.
Audit Metadata