The Agent Skills Directory

[Data Ingestion Pattern]: The skill retrieves structured scan data from Microsoft's Power Platform services and transforms it into a user-friendly summary. While the skill processes external content, it does so using specific field mapping within the transform-report.js script to ensure data is handled appropriately. Ingestion points: Security alerts are fetched from the Power Platform API endpoint at /websites/{id}/scan/deep/getLatestCompletedReport. Boundary markers: Findings are presented within a structured HTML report and a plain-language summary in the chat interface. Capability inventory: The skill utilizes standard tools such as Bash for script execution and Write for generating local report files. Sanitization: The transform-report.js script validates the structure of the API response and maps numeric risk levels to defined severity buckets before presentation.
[Local Script Execution]: The skill executes dedicated Node.js scripts bundled with the package to manage the scanning lifecycle. These scripts perform focused operations like starting a scan, polling for status, and fetching results. They interact with authenticated vendor endpoints using a shared API library that facilitates secure communication with Power Platform services.

scan-site