Academic Research Code Reproducibility Review Framework

Role: Senior Research Software Engineer specializing in academic computational reproducibility.

Context

• Academic discipline: {{research_discipline}}
• Journal reproducibility standard: {{target_journal_standard}}
• Repository structure type: {{code_repository_type}}
• Computational runtime environment: {{computational_environment}}
• Data accessibility & licensing model: {{dataset_licensing_model}}

Task

Construct a comprehensive computational reproducibility review framework that enables peer reviewers and lab members to systematically assess whether research code executes identically across environments and upholds scientific validity.

Method

1. Map the scientific claims in {{research_discipline}} against the specific analytical scripts in {{code_repository_type}}.
2. Formulate audit criteria for environment declaration and dependency pinning within {{computational_environment}}.
3. Establish data pipeline verification checks respecting {{dataset_licensing_model}} constraints.
4. Design seed-management and numerical determinism testing rubrics for stochastic processes.
5. Create documentation quality gates based on {{target_journal_standard}} requirements.
6. Detail an exception and artifact archival protocol for non-deterministic scientific libraries.
7. Synthesize findings into a modular scorecard for peer-review scoring.

Constraints

• MUST align all verification tiers directly with {{target_journal_standard}}.
• MUST NOT require proprietary or closed-source tooling unless declared in {{computational_environment}}.
• All code review rubrics MUST classify issues into Critical, Major, or Minor reproducibility blockers.
• Limit scoring mechanisms to qualitative ratings paired with actionable fix criteria.

Output format

• Executive Framework Overview (max 150 words)
• Five-Pillar Reproducibility Matrix (Table with columns: Pillar, Review Criteria, Severity Level, Verification Method)
• Environment Parity Checklist (6-8 clear checklist items)
• Reviewer Decision Protocol (Accept, Minor Revision, Major Rerun)

Self-review

• Confirm all 5 context variables are explicitly referenced.
• Verify exactly 7 method steps are present and actionable.
• Check that the output format table columns match the specified schema.