Step 8: Export

Purpose

The Export step sends the final grading score to an external system. This is used in contexts where the autograder needs to report results outside of its own database — for example, updating a student's score in a Redis-backed leaderboard or quota system.

How It Works

1. Extract data — The step reads the external_user_id from the submission and the final_score from the Grade step result.
2. Send to external system — The exporter_service.set_score(user_id, score) call pushes the score to the configured backend.

Currently, the only implemented exporter is UpstashDriver, which stores scores in an Upstash Redis instance using the hash key user:{username} with a score field.

Dependencies

Step	What It Needs
Grade	The final_score from GradeStepResult

Input

Source	Data
Constructor	exporter_service — the service that handles external communication
Pipeline	StepName.GRADE → GradeStepResult.final_score
Pipeline	pipeline_exec.submission.user_id

Output

Field	Type	Description
data	None	No data is stored — the export is a side effect
status	StepStatus.SUCCESS	On successful export

Conditional Inclusion

This step is only added to the pipeline when export_results=True is passed to build_pipeline(). It is typically used in the GitHub Action workflow where scores need to be reported to GitHub Classroom via an intermediary store.

Failure Scenarios

• External service unavailable (Redis connection error) → StepStatus.FAIL with error message.
• Missing environment variables for Upstash (UPSTASH_REDIS_URL, UPSTASH_REDIS_TOKEN) → connection failure.

Export failures do not affect the grading result — the score and feedback are already computed. However, the pipeline status will reflect the failure.

---

Source

autograder/steps/export_step.py → ExporterStep

autograder/services/upstash_driver.py → UpstashDriver