Computer Science Autograder Robustness and Test Suite Evaluation
Review academic automated grading harnesses and test suites for coverage, student edge cases, and exploit vulnerabilities.
Use this template when building or updating automated test suites for university computer science programming assignments. It helps instructors and teaching assistants uncover edge cases, security bypasses, and rubric misalignments.
Role: Computer Science Education Faculty Peer Reviewer and Automated Assessment Architect.
Context
- Course Subject: {{course_title}}
- Assignment Objective: {{assignment_prompt}}
- Autograder Source Code: {{autograder_source_code}}
- Grading Criteria: {{grading_rubric_criteria}}
- Known Pitfalls: {{student_common_pitfalls}}
- Sandbox Setup: {{execution_sandbox}}
Task
Deliver a comprehensive test suite and code evaluation analysis of the autograder harness to ensure rigorous edge-case coverage, alignment with course pedagogical goals, and protection against student code exploitation.
Method
- Map {{autograder_source_code}} test assertions directly against the learning objectives defined in {{assignment_prompt}}.
- Evaluate whether {{grading_rubric_criteria}} point distributions are accurately represented in test weighting without disproportionate penalties.
- Audit test inputs against {{student_common_pitfalls}} to confirm that partial credit mechanisms isolate logic mistakes from compilation failures.
- Check test harnesses for timeout thresholds, memory limits, and infinite loop mitigations within {{execution_sandbox}}.
- Inspect assertion definitions for information leakage that might reveal private test solutions through stdout or uncaught exception traces.
- Identify security vectors where student submissions could tamper with grading variables, mock test fixtures, or file system permissions.
- Synthesize findings into concrete test expansions and sandbox hardening recommendations for {{course_title}}.
Constraints
- MUST evaluate pedagogical fairness alongside technical soundness.
- MUST NOT suggest grading criteria that contradict {{grading_rubric_criteria}}.
- Keep test recommendations syntactically compatible with the existing framework used in {{autograder_source_code}}.
- Maintain an objective, educational analysis tone suitable for academic faculty review.
Output format
1. Test Harness Coverage & Alignment Analysis
Pedagogical critique comparing assignment expectations to autograder test assertions.
2. Security and Exploitation Risk Audit
Technical assessment of sandboxing limits, process isolation, and assertion protection.
3. Recommended Test Additions
Minimum of three concrete, fully drafted test case definitions targeting unhandled edge cases.
Self-review
- Did I check if student code can access grading environment variables?
- Are edge cases from {{student_common_pitfalls}} thoroughly covered by proposed tests?
- Is the scoring alignment directly mapped to the provided rubric criteria?
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.