review
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from source code files, pull request diffs, and project-specific documentation. This ingestion surface is susceptible to indirect prompt injection where malicious instructions embedded in the code could influence the agent. 1. Ingestion points: Source code files (.tsx), PR diffs, and configuration files (CLAUDE.md). 2. Boundary markers: The instructions lack explicit delimiters or safety headers for the analyzed content. 3. Capability inventory: File system access, tool execution (axe, Lighthouse), and dynamic loading of modular rulesets. 4. Sanitization: The skill does not implement sanitization for the data it processes.
- [COMMAND_EXECUTION]: The skill instructs the agent to run the interface locally and utilize external CLI tools for auditing, which requires process execution capabilities within the local environment.
Audit Metadata