The Agent Skills Directory

Execution of Local Scripts: The skill executes several JavaScript files (check-version.js, check-activation-status.js, and refresh-alm-plan-data.js) located within the plugin's own directory to verify versions and project status.
External Data Ingestion: The skill crawls content from a target website, including links, page text, and API responses. Processing this untrusted external data presents an indirect prompt injection surface, which is managed here by a page crawl limit and focused diagnostic intent. Ingestion points: Website content and API responses. Boundary markers: None explicitly used. Capability inventory: Includes shell access, file system reads, and extensive browser automation tools. Sanitization: Link extraction logic filters for same-origin URLs.
Dynamic File Operations: It uses shell commands to dynamically create directories and write the final test report to the local file system.
Browser Code Execution: The skill uses browser_evaluate to execute specific JavaScript snippets within the browser context to extract internal links and analyze API response structures.
Authentication Best Practices: The skill follows safe practices by requiring the user to perform manual authentication through the browser window rather than attempting to automate credential entry.

test-site