skills/garethmanning/education-agent-skills/cognitive-tutoring-architecture-designer/Gen Agent Trust Hub
cognitive-tutoring-architecture-designer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data by interpolating several input variables into its core instructions without sufficient isolation.
- Ingestion points: The input variables
{{skill_domain}},{{knowledge_components}},{{student_level}},{{subject_area}},{{common_errors}},{{mastery_threshold}}, and{{system_context}}are used in thePromptsection ofSKILL.md. - Boundary markers: The skill lacks explicit delimiters or specific instructions to the model to ignore any control sequences or instructions embedded within the user-provided content.
- Capability inventory: No dangerous operations like file system writes, network requests, or shell command executions were found in the skill's instructions.
- Sanitization: There is no evidence of input sanitization or validation mechanisms to prevent adversarial input from influencing the agent's behavior.
Audit Metadata