backwards-design-unit-planner
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a non-executable, prompt-only configuration. It does not contain scripts, binaries, or instructions that interact with the host filesystem or external network.
- [REMOTE_CODE_EXECUTION]: No remote dependencies or script downloads were identified. All logic is contained within the static Markdown prompt.
- [DATA_EXFILTRATION]: The skill does not perform any network operations. It does not access sensitive local paths (e.g., .ssh, .aws) or environment variables.
- [PROMPT_INJECTION]: The internal prompt is focused on educational design. There are no instructions to bypass safety guidelines, reveal system prompts, or disregard previous constraints.
- [PROMPT_INJECTION]: (Indirect Surface) The skill processes user-supplied curriculum outcomes through template variables (e.g., {{desired_outcomes}}). While the prompt lacks explicit delimiters or sanitization for these inputs, the risk is negligible because the skill does not have access to any tools or capabilities (file writing, shell execution, network) that could be leveraged in an injection attack.
Audit Metadata