llm-application-dev-ai-assistant
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates untrusted user input via the $ARGUMENTS variable in the Requirements section of both SKILL.md and resources/implementation-playbook.md without clear boundary markers or instructions to disregard embedded commands.\n
- Ingestion points: Untrusted data enters the agent context via the $ARGUMENTS placeholder in the Requirements sections of SKILL.md and resources/implementation-playbook.md.\n
- Boundary markers: There are no explicit delimiters (e.g., XML tags) or "ignore embedded instructions" warnings surrounding the interpolation of user-supplied requirements.\n
- Capability inventory: The skill enables the agent to read internal resource files and provides code templates for building components with network access (LLM integration layer) and deployment automation (Docker/Kubernetes).\n
- Sanitization: No sanitization, escaping, or validation of the user-supplied content is performed before it is interpolated into the agent's context.
Audit Metadata