research-gap-analyst - SKILL.md Agent Skill

name: research-gap-analyst description: Evaluates the codebase against Scopus-indexing standards. It should identify missing experiments, hyperparameter details, or validation metrics. Save to docs/paper/analysis/gap_report.md.

This skill critiques the project from a research perspective, identifying gaps that need to be addressed to meet Scopus-indexing standards.

Analyze Codebase: Review the code, documentation, and existing results.
Identify Gaps: Look for:
- Missing experiments or validation scenarios.
- Undefined or hardcoded hyperparameters.
- Lack of baseline comparisons.
- Insufficient performance metrics.
Verify Source: strict verification against the actual source code is required. Do not assume something is missing without checking the entire codebase.
Output:
- File: docs/paper/analysis/gap_report.md
- Format: Markdown report.
- Content:
  - List of missing experiments.
  - Details on missing hyperparameter explanations.
  - Suggestions for improving the "scientific rigor" of the project.

Double-check that identified "missing" features are indeed strictly absent from the codebase.

gap_checklist.md - Comprehensive checklist covering experimental rigor, hyperparameter documentation, performance validation, reproducibility, code quality, and academic standards with gap severity classification

gap_analysis_example.md - Complete gap analysis for the Amsha project showing critical, moderate, and minor gaps with specific file locations, impact assessments, recommendations, and effort estimates

When using these materials:

Checklist: Use all 6 sections systematically (experimental, hyperparameters, performance, reproducibility, code quality, academic)
Severity: Critical = blocks publication, Moderate = weakens paper, Minor = nice-to-have
Recommendations: Must be specific and actionable with effort estimates
Verification: Every gap must reference actual code location (file:line)