The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the @useclawlink/cli via npx. These commands include login, connect, actions, describe, and run, which are necessary for the skill's functionality.
[EXTERNAL_DOWNLOADS]: The skill uses npx @useclawlink/cli to download and run the ClawLink CLI from the official NPM registry. This package is provided by the skill's author (clawlink-hq) and is a standard delivery method for this tool.
[DATA_EXPOSURE]: The skill documentation notes that credentials are stored locally at ~/.clawlink/credentials.json. This is standard behavior for the CLI tool to manage user authentication state across sessions and does not involve exfiltration to unauthorized third parties.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external data from Google Classroom (e.g., student submissions and coursework content).
Ingestion points: Untrusted data enters the agent context through actions such as google_classroom_course_work_student_submissions_list and google_classroom_course_work_list in SKILL.md.
Boundary markers: None explicitly defined in the provided instructions; the agent relies on its internal reasoning to distinguish data from instructions.
Capability inventory: The skill has the capability to execute shell commands via npx and read/write data to Google Classroom via the ClawLink CLI.
Sanitization: No specific sanitization or filtering of student submissions or coursework content is mentioned.

google-classroom