The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from existing public-class teaching sites, which constitutes an untrusted data surface.
Ingestion points: Existing course units, material files, and site content from project-root/ are read to be condensed and repackaged.
Boundary markers: There are no instructions for the agent to use delimiters or ignore potentially malicious instructions embedded within the source course units.
Capability inventory: The skill performs file system operations (reading/writing index.html, materials/, assets/) and executes JavaScript-based verification scripts.
Sanitization: The instructions do not specify any validation or sanitization of the input text before it is inlined into the window.COURSE object in the final HTML deliverable.
[COMMAND_EXECUTION]: The workflow relies on the execution of multiple companion scripts to verify the integrity of the generated corporate edition.
Evidence: The skill references scripts such as scripts/verify-corp-{name}.mjs and scripts/audit-corp-{name}-content.mjs. These scripts utilize tools like Playwright to perform DOM inspections and screenshotting, which involves executing code and interacting with the local system environment.

course-corporate-edition