The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill demonstrates best practices in security orchestration and developer tooling.\n- [EXTERNAL_DOWNLOADS]: The skill utilizes the well-known and trusted Playwright library (maintained by Microsoft) for rendering JavaScript-heavy websites during the audit process.\n- [COMMAND_EXECUTION]: The orchestration logic executes several internal Node.js and Python scripts (e.g., analyze.mjs, fetch.mjs, compare.mjs) located within the plugin's own toolkit directory to perform deterministic checks and manage audit data.\n- [PROMPT_INJECTION]: The skill proactively mitigates indirect prompt injection by including a 'Trust boundary' directive, ensuring that content fetched from external URLs is treated strictly as data and not as instructions for the agent.

audit