bounded-context-mapping
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill. The instructions define a conceptual methodology for architectural mapping.
- [PROMPT_INJECTION]: The skill includes instructions to extract and analyze terms from external sources such as requirements, code, and documentation. Although this constitutes an ingestion surface for untrusted data, the skill is limited to read-only operations (
Read,Grep) and lacks any execution or network capabilities that would enable an exploit through indirect injection. - Ingestion points: Method section describes extracting terms from requirements, code, and docs.
- Boundary markers: None specified.
- Capability inventory: Restricted to
ReadandGreptools. - Sanitization: Not applicable as no execution occurs.
Audit Metadata