narrative-designer
Pass
Audited by Gen Agent Trust Hub on May 1, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
!catsyntax inSKILL.mdto include external documentation and configuration files (e.g.,skills/_shared/game-visual-foundations.md,.production-grade.yaml) into the agent's context at load time. These commands are restricted to reading local files within the expected directory structure and do not incorporate unvalidated user input.\n- [PROMPT_INJECTION]: The skill processes untrusted data from game design documents.\n - Ingestion points:
SKILL.md(GDD, level output, mechanics inputs)\n - Boundary markers: Absent\n
- Capability inventory: No dangerous tools or subprocess operations identified\n
- Sanitization: Absent
Audit Metadata