research-engineer

name	research-engineer
description	Structured technical/theoretical research workflow for method selection, model comparison, and implementation-oriented conclusions with validation plans. Use when decisions must map literature evidence to concrete engineering or computational next steps.
license	MIT
metadata	{"author":"local-adapted","version":"1.0.0"}

Research Engineer

When to apply

• Need to compare competing methods or models.
• Need evidence-backed recommendations for implementation choices.
• Need to turn literature or benchmark evidence into concrete engineering or computational next steps.

Boundary

• Broad background gathering, perspective mapping, or open-ended topic familiarization goes to deep-research.
• Reproducible screening-oriented review workflows go to literature-review.
• Reference cleanup, DOI normalization, or BibTeX curation goes to citation-management.
• Use this skill when the end product is a decision, plan, or implementation path rather than a survey alone.

Research protocol

1. Define scope: domain, assumptions, constraints, and success criteria.
2. Build search plan: keywords, variants, and exclusion terms.
3. Collect sources: prioritize peer-reviewed and authoritative materials.
4. Evaluate quality: methodology, reproducibility, date, and relevance.
5. Synthesize: consensus, disagreement, gaps, and practical implications.

Evidence standards

• Separate established facts from hypotheses.
• Mark uncertain claims explicitly.
• Prefer recent sources when the field evolves rapidly.
• Highlight reproducibility signals (code/data availability, benchmark quality).

Output template

Executive Summary

• 3-6 bullet conclusions with confidence levels.

Method Comparison

• Assumptions
• Strengths
• Limitations
• Computational cost
• Suitability for this project

Recommended Path

• Short-term actionable steps
• Validation plan
• Risks and fallback options

Source Notes

• Numbered source list with one-line credibility notes

Hand-off

• Upstream inputs often come from deep-research or literature-review.
• Downstream outputs should usually feed implementation work, experiments, or benchmark validation.
• If the task ends at bibliography hygiene rather than method choice, hand off to citation-management.