ddd4j-boot-web
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill manages web request lifecycles, which inherently involves processing untrusted external data via Spring MVC or WebFlux controllers. The instructions focus on secure implementation patterns for these ingestion points.
- Ingestion points: HTTP requests and web-layer interactions processed by the ddd4j-boot modules (SKILL.md).
- Boundary markers: The skill emphasizes the use of unified error contracts and fictional data for request/response examples to prevent leakage.
- Capability inventory: Capabilities are limited to configuring framework-level beans for CORS, rate limiting, and health checks. There are no capabilities for arbitrary command execution or unauthorized file system writes.
- Sanitization: Instructions explicitly mandate the redaction of sensitive configuration values, such as hosts and tokens, when debugging (SKILL.md).
- [EXTERNAL_DOWNLOADS]: The documentation includes references to the
npx skills add full-stack-skills/ddd4j-skillscommand. This is a standard platform mechanism for adding related skills from the same vendor ('full-stack-skills') and does not represent an untrusted source.
Audit Metadata