The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes @useclawlink/cli, which is downloaded and executed via npx from the npm registry. This is a vendor-owned resource managed by clawlink-hq and is used for its intended purpose of providing the interface between the agent and the ClawLink service.
[COMMAND_EXECUTION]: The skill functions by executing shell commands to interact with the Google Classroom API via the ClawLink CLI. This includes commands for logging in, connecting services, and running specific classroom actions.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it retrieves content from Google Classroom that may be created by students or other teachers.
Ingestion points: SKILL.md defines actions such as google_classroom_course_work_student_submissions_list and google_classroom_course_work_list which fetch untrusted data into the agent's context.
Boundary markers: The instructions include a clear directive: "Reads first; confirm with the user before any write," which serves as a human-in-the-loop safety boundary.
Capability inventory: The skill provides the ability to modify Classroom data (create/delete coursework) using the npx @useclawlink/cli run command.
Sanitization: No automated sanitization of retrieved text content is specified, relying instead on the user confirmation step.

google-classroom