duo-measurement-validity
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The analyzed content consists strictly of informational Markdown files describing best practices for data science and product measurement. No security threats were identified.
- [NO_CODE]: The skill does not include any executable scripts, shell commands, or configuration files that would permit code execution. All logic is provided as natural language instructions for the agent to follow when assisting with measurement design.
- [METADATA_POISONING]: The YAML frontmatter and skill metadata fields contain standard identification information (name, author, version) and do not include any hidden or deceptive instructions.
- [EXTERNAL_DOWNLOADS]: While the documentation references external sources such as 'blog.duolingo.com' for context and citations, there are no commands to download or execute remote payloads.
Audit Metadata