wcag-audit-patterns
Warn
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to install tools and patterns from a third-party GitHub repository (wshobson/agents) using the
npx skills addcommand and marketplace plugin installation commands. - [COMMAND_EXECUTION]: The installation instructions include shell scripts that execute logic to detect existing files and grep for specific strings in the local filesystem to determine the installation state.
- [INDIRECT_PROMPT_INJECTION]: The skill performs audits on web interfaces, creating a surface for instructions embedded in the analyzed content to influence the agent.
- Ingestion points: Analyzes web interfaces, including HTML, CSS, and other frontend assets (SKILL.md).
- Boundary markers: No specific boundary markers or instructions to treat analyzed content as untrusted data are provided.
- Capability inventory: Uses analyzed data to generate remediation guidance and prioritized checklists.
- Sanitization: No evidence of content sanitization or validation is present in the skill definition.
Audit Metadata