audit-accessibility
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data by scraping WCAG guidelines and framework documentation via Firecrawl. * Ingestion points: External URLs scraped during Phase 1. * Boundary markers: None identified; external content is provided to the agent context without explicit delimiters. * Capability inventory: The skill possesses capabilities including local file system access, network search/scrape, and browser automation with script execution. * Sanitization: No evidence of sanitization or instructions to ignore embedded commands in the external data.- [DYNAMIC_EXECUTION]: The skill utilizes the Playwright CLI's run-code and eval features to execute JavaScript logic within a browser instance to perform accessibility checks.- [COMMAND_EXECUTION]: The skill automates several shell-based tasks, including package execution via npx and local file system discovery using Glob and Grep.- [EXTERNAL_DOWNLOADS]: The skill fetches the axe-core library from cdnjs.cloudflare.com and installs the Playwright CLI using npx. These are well-known resources used for their intended purpose in an accessibility audit.
Audit Metadata