The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill consists of natural language instructions focused on accessibility improvements, such as ARIA labels, keyboard navigation, and WCAG compliance.
[NO_CODE]: The skill does not include any scripts, binary executables, or external software dependencies, which significantly limits the potential attack surface.
[INDIRECT_PROMPT_INJECTION]: While the skill processes external files to audit for accessibility violations, it lacks dangerous capabilities like network access, file writing, or command execution, meaning that any embedded instructions in the audited data cannot trigger malicious actions.

fixing-accessibility