메뉴
Use when an experiment phase is complete, before drawing conclusions, before writing results, or before claiming findings are publication-ready
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use during or after manuscript writing to verify every quantitative claim traces to actual data, every figure is script-generated, and all results including negative ones are reported
Use when starting any conversation involving research, hypotheses, experiments, data analysis, scientific claims, or manuscript writing — establishes how to find and use Eureka research skills
Use when a research design has been approved and needs to be broken into executable experiment steps — the research equivalent of an implementation plan
Use when creating, updating, or reviewing a research figure (plot, chart, brain map, heatmap, schematic) — guides chart-type selection, typography, colorblind-safe palette, layout, and journal-specific export specs. Dispatches a figure-reviewer subagent after rendering. Complements claims-audit (which checks figure integrity) by checking figure design.
Use when running any experiment or analysis — hypothesis and analysis plan must be registered before seeing results, the scientific equivalent of test-driven development
Use when writing a research manuscript — guides section-by-section writing with prerequisite gates, citation discipline, number traceability, and per-section subagent review. Activated after experiment results are reviewed and before claims-audit.
| name | requesting-research-review |
| description | Use when an experiment phase is complete, before drawing conclusions, before writing results, or before claiming findings are publication-ready |
Dispatch eureka:research-reviewer subagent to catch scientific issues before they propagate. The reviewer gets precisely crafted context for evaluation — never your session's history. This keeps the reviewer focused on the evidence, not your thought process, and preserves your own context for continued work.
Core principle: Review before you conclude.
Mandatory:
Optional but valuable:
1. Identify what you're reviewing:
Determine which research phase just completed and what the reviewer needs to evaluate.
2. Gather the review context:
RESEARCH_QUESTION = The exact hypothesis being tested
DESIGN_DOC_PATH = Path to the approved research design document
RESULTS_PATHS = Paths to ALL result files (not just summaries)
PHASE_DESCRIPTION = Which phase just completed (e.g., "baseline experiments")
DOMAIN_CONTEXT = Field-specific context, tools, typical effect sizes
TARGET_VENUE = Target journal or conference (sets the bar)
PASS_THRESHOLD = Score threshold (85 for mid-project, 95 for pre-submission)
3. Dispatch the research-reviewer subagent:
Use the Agent tool with eureka:research-reviewer type. Fill the template at requesting-research-review/research-reviewer-prompt.md with the gathered context.
4. Act on feedback:
| Situation | Threshold | Reasoning |
|---|---|---|
| Mid-project checkpoint | 85/100 | Catch major issues early without demanding final polish |
| Before writing results | 90/100 | Claims must be well-supported before committing to paper |
| Pre-submission gate | 95/100 | Publication-ready: all claims verified, all stats complete |
| Quick sanity check | 80/100 | Just checking for obvious problems |
[Just completed baseline experiments comparing three models on a held-out test set]
You: Let me request a research review before interpreting these results.
RESEARCH_QUESTION = "Does Model C outperform Model A and Model B baselines
on outcome Y prediction?"
DESIGN_DOC_PATH = docs/eureka/designs/2026-04-10-model-baselines-design.md
RESULTS_PATHS = results/run_20260411_171136/cv_results.json,
results/baseline_modelA/, results/baseline_modelB/
PHASE_DESCRIPTION = Baseline experiment completion (Model A, Model B, Model C)
DOMAIN_CONTEXT = Supervised prediction on cohort data,
Pearson r and R-squared are standard metrics
TARGET_VENUE = [target journal]
PASS_THRESHOLD = 85
[Dispatch eureka:research-reviewer subagent]
[Subagent returns]:
D1: 88 D2: 82 D3: 75 D4: 70 D5: 72 D6: 78 D7: 80
Overall: FAIL (D3, D4 below threshold)
Critical: Single seed results reported — need ≥5 seeds
Major: No confidence intervals on Pearson r comparisons
You: [Fix: re-run with 10 seeds, add bootstrap CIs]
[Request re-review after fixes]
eureka:using-eureka (when experiment phase completion detected)eureka:research-reviewer agent (via Agent tool)eureka:receiving-research-review (for handling the feedback)Never:
If reviewer is wrong:
See template at: requesting-research-review/research-reviewer-prompt.md