a11y-checker

Warn

Audited by Gen Agent Trust Hub on Oct 3, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to run npx ai-a11y, which downloads and executes a package from the public NPM registry. The package ai-a11y and its maintainer LXGIC Studios are not verified trusted sources.
  • [COMMAND_EXECUTION]: The core functionality of the skill is provided through shell command execution using npx, allowing external code to run on the host system.
  • [REMOTE_CODE_EXECUTION]: Running a package directly from NPM via npx executes code managed by an external third party, which could be compromised or malicious.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes HTML and JSX files to generate AI-driven accessibility fixes, which presents a vulnerability to malicious instructions embedded in the scanned files.
  • Ingestion points: Source code files provided as arguments to the tool (e.g., ./src/components/Header.tsx as referenced in SKILL.md).
  • Boundary markers: None identified in the instructions to prevent the AI from interpreting content within the scanned files as commands.
  • Capability inventory: Shell command execution capability via npx documented in SKILL.md.
  • Sanitization: There is no indication that the content of the files being scanned is sanitized or escaped before being processed by the AI component.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 3, 2026, 11:29 PM