accesibilidad-wcag
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The scripts
audit-axe.sh,a11y-check.mjs, andcontrast-check.mjsare designed to accept sensitive passwords as command-line arguments. This practice can expose secrets in process lists, shell history, and logs. - [COMMAND_EXECUTION]: In
scripts/audit-axe.sh, shell variables$A11Y_USERand$A11Y_PASSare interpolated directly into a Javascript string that is then executed viaagent-browser eval. This lacks proper sanitization; if the credentials contain single quotes or other control characters, they could escape the string literal and execute arbitrary Javascript within the browser context. - [INDIRECT_PROMPT_INJECTION]: The skill interacts with external, untrusted web content to perform audits, which represents a potential attack surface.
- Ingestion points: The
agent-browserand Playwright scripts fetch and process data from arbitrary URLs provided as input. - Boundary markers: None identified; the skill assumes the content of the audited pages is data rather than instructions.
- Capability inventory: The skill has capabilities to execute shell commands, automate browser actions, and write JSON reports to the local file system.
- Sanitization: The skill relies on standard library tools (axe-core, Playwright) to process page content, which generally treat DOM elements as data, but the
evallogic in the shell script presents a path where data could influence execution.
Audit Metadata