fullstack-guardian
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is designed to improve the security posture of generated code. It provides detailed references for authentication, authorization, and error handling, while explicitly forbidding unsafe practices such as hardcoding credentials or trusting client-side validation alone.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external feature requirements to generate application code, which presents a surface for indirect prompt injection. However, it incorporates robust defensive measures to handle untrusted input safely.
- Ingestion points: Feature requirements and acceptance criteria are processed during the core implementation workflow defined in
SKILL.md. - Boundary markers: The workflow requires the creation of technical design documents (
specs/{feature}_design.md) and the use of a security checklist (references/security-checklist.md) to verify the scope and safety of the implementation. - Capability inventory: The skill allows the agent to generate backend (FastAPI, NestJS) and frontend (React) code, including database migrations, API endpoints, and Docker configurations across all referenced files.
- Sanitization: The instructions mandate server-side validation using schemas (Zod/Pydantic), output sanitization to prevent XSS, and the use of parameterized queries for all database interactions.
Audit Metadata