architecture-strategist
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it is designed to read and analyze external, potentially untrusted data from the local filesystem.
- Ingestion points: The skill reads architecture documentation, README files, and source code changes (SKILL.md).
- Boundary markers: There are no boundary markers or explicit instructions provided to the agent to ignore or delimit potentially malicious instructions embedded within the files it analyzes.
- Capability inventory: The skill's primary function is to read files to map dependencies and analyze system design; it does not explicitly specify network or file-writing capabilities in its instructions.
- Sanitization: The instructions do not include requirements for sanitizing or validating the content extracted from project files before processing.
Audit Metadata