skills/garethmanning/claude-education-skills/experiential-learning-cycle-designer/Gen Agent Trust Hub
experiential-learning-cycle-designer
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely content-oriented and does not utilize any tools, file system operations, or network requests.
- [PROMPT_INJECTION]: Instructions within the prompt are focused on educational methodology (e.g., structuring reflection and conceptualization). There are no attempts to bypass safety filters, extract system prompts, or override agent behavior.
- [DATA_EXPOSURE]: The skill does not access sensitive files, environment variables, or credentials. It operates only on user-provided educational parameters.
- [REMOTE_CODE_EXECUTION]: No external dependencies, package installations, or remote script execution patterns are present in the skill definition.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests user-provided text (such as learning objectives) into a prompt template, it lacks any execution capabilities (subprocess, file-write, network) that would allow such an injection to have a malicious impact beyond text generation.
Audit Metadata