practice-problem-sequence-designer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates entirely within the prompt context to generate educational text. It does not request or use external tools, perform network operations, or access sensitive local files.- [INDIRECT_PROMPT_INJECTION]: The skill accepts multiple user-controlled input fields which are interpolated into the final prompt, creating a potential surface for indirect injection.
- Ingestion points: Fields such as
skill_to_practise,student_level, andprior_examplesinSKILL.md. - Boundary markers: The prompt uses bold markdown headers as delimiters, though it lacks explicit instructions for the AI to ignore embedded commands within the inputs.
- Capability inventory: No tool invocation, shell command execution, or network access capabilities are defined in the skill configuration.
- Sanitization: No sanitization or validation of the input strings is performed.
- Conclusion: Due to the absence of dangerous capabilities, the risk is negligible and limited to potential manipulation of the generated text output.
Audit Metadata