跳到主要内容
知仓学习社ZHICANG

paper-excellence

Comprehensive multi-dimensional review of the sewage-house-prices project. Runs econometrics audit, code review, manuscript proofread, and bibliogra…

不碰外部(只输出文字)无严重或高危命中brycewang-stanford/Auto-Empirical-Research-Skills

它会碰到什么

扫了多少1 个文本文件,4 KB
它会碰到什么不碰外部(只输出文字)
命中总数0 处
命中统计严重 0 · 高 0 · 中 0 · 低 0

这一栏是扫描器报的事实,不是结论。命中多不等于有毒(安全工具、规则库、示例脚本本来就会包含危险写法),命中少也不等于干净。它和你手上的凭据、文件、网络有什么关系,需要你自己看。

技能内容

Paper Excellence Review

Run a comprehensive quality assessment of the sewage-house-prices project across all dimensions.

Input: $ARGUMENTSall for full review, or a specific component.


Workflow

Step 1: Identify Targets

  • docs/overleaf/*.tex — Manuscript sections
  • scripts/R/09_analysis/ — Analysis scripts
  • output/tables/ — Generated tables
  • output/figures/ — Generated figures
  • docs/overleaf/refs.bib — Bibliography

Step 2: Launch Review Agents (Parallel)

Launch up to 4 agents simultaneously:

Agent 1: Econometrics Audit

Review all identification strategies (hedonic, repeat sales, long diff, DiD, upstream/downstream, dry spills).

Cross-reference manuscript claims against analysis scripts.

Weight: 30%

Agent 2: Code Review

Review all scripts in scripts/R/09_analysis/ for code quality, reproducibility, and project convention compliance.

Weight: 15%

Agent 3: Manuscript Proofread

Review all .tex files for structure, claims-evidence alignment, identification fidelity, writing quality, grammar, and LaTeX compilation.

Weight: 35%

Agent 4: Bibliography Validation

Cross-reference all citations against refs.bib. Check for missing entries, unused references, and quality issues.

Weight: 5%

Step 3: Compute Weighted Aggregate Score

Overall = 0.30 × Econometrics + 0.15 × Code + 0.35 × Paper + 0.05 × Bibliography + 0.15 × Polish

Where Polish is derived from the Proofreader's writing quality subscore.

If components are missing (e.g. no manuscript sections yet), renormalise weights over available components.

Step 4: Present Results

# Paper Excellence Report: Sewage in Our Waters
**Date:** YYYY-MM-DD
**Aggregate Score:** XX/100

## Score Breakdown
| Component | Weight | Score | Issues | Source |
|-----------|--------|-------|--------|--------|
| Econometrics | 30% | XX | N | /econometrics-check |
| Code | 15% | XX | N | /review-r |
| Paper | 35% | XX | N | /proofread |
| Bibliography | 5% | XX | N | /validate-bib |
| Polish | 15% | XX | N | Writing quality subscore |

## Priority Fixes (Top 5)
1. **[CRITICAL]** [Most important issue]
2. **[MAJOR]** [Second priority]
3. ...

## Quality Gate
- Score >= 90: "Ready for submission."
- Score >= 80: "Commit-ready. Address major issues before submission."
- Score < 80: "Blocked. Must fix critical/major issues."

## Full Reports
- Econometrics: output/log/econometrics_check_all.md
- Code: output/log/code_review_all.md
- Proofread: output/log/proofread_report_all.md
- Bibliography: output/log/bib_validation.md

Save to output/log/paper_excellence_[date].md.


Principles

  • Parallel execution. All agents run simultaneously for efficiency.
  • Weighted aggregation. Not a simple average — econometrics and paper quality dominate.
  • Don't double-count. Same issue found by multiple agents counts once in priority list.
  • One unified report. User sees one priority list, not separate reports.
  • Proportional gating. Working papers get developmental feedback. Near-final manuscripts get submission-level scrutiny.

想直接用这个技能?

本站把开放许可(MIT / Apache 等)的技能按仓库打包整理到网盘,点一下转存到你自己的网盘,不用一个个从 GitHub 拉。许可未声明的技能只给原始仓库链接,不打包。