What problem does it solve? Instructors running Canvas peer reviews need pairing to happen automatically at a locked deadline without manual runs, while keeping student data private and credentials out of the course repository. ## Core Features & Use Cases - Scheduled pairing workflow: Installs a course-owned GitHub Actions workflow that polls on a cron cadence and gates execution to a configured local-time window using zoneinfo, so DST changes require no edits. - Dry-run safety by default: The scheduler runs as a no-op outside the configured window or before the lock time, and only writes pairings after explicit validation against a Canvas sandbox. - Secret-safe setup: A helper script uploads only CANVAS_API_TOKEN and CANVAS_BASE_URL as repository secrets and the course ID as a variable, never printing secret values. - Use Case: An instructor configures a peer-review schedule for a group assignment, validates the workflow manually with apply unchecked, then enables the schedule so pairings are assigned automatically after the submission lock time. ## Quick Start Ask the agent to set up the scheduled peer-review workflow for your course repository and walk through validating it in dry-run mode against a Canvas sandbox.
