菜单
Multi-agent paper review dispatching econometrics-critic, debugger, academic-proofreader, and replication-verifier in parallel. Computes weighted aggregate score per scoring-protocol.md. Use for comprehensive quality check before submission or major milestones.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | paper-excellence |
| description | Multi-agent paper review dispatching econometrics-critic, debugger, academic-proofreader, and replication-verifier in parallel. Computes weighted aggregate score per scoring-protocol.md. Use for comprehensive quality check before submission or major milestones. |
| disable-model-invocation | true |
| argument-hint | [paper .tex path OR 'all' for full project review] |
| allowed-tools | ["Read","Grep","Glob","Write","Task","Bash"] |
Run all review agents in parallel for a comprehensive paper quality assessment. Computes the weighted aggregate score from scoring-protocol.md.
Input: $ARGUMENTS — path to paper .tex file, or all for full project review.
$ARGUMENTS is a .tex file: review that file$ARGUMENTS is all: review paper/main.tex + all scripts in scripts/R/Talks/*.tex for auxiliary scoringBefore launching agents:
.claude/rules/domain-profile.md for field-specific calibration.claude/rules/scoring-protocol.md for weight formulaquality_reports/Launch up to 4 agents simultaneously via Task tool:
Agent 1: econometrics-critic (subagent_type: econometrics-critic)
Review [paper.tex] through all 4 phases: claim, design validity, inference, polish.
Also check scripts in scripts/R/ for code-theory alignment.
Save report to quality_reports/[file]_econometrics_review.md
Weight: Identification 25% of aggregate score.
Agent 2: Debugger (subagent_type: general-purpose, with debugger agent instructions)
Review all scripts in scripts/R/ for code quality and correctness.
Run categories 4-12 (code quality) plus categories 1-3 (strategic) if strategy memo exists.
Save report to quality_reports/[script]_code_review.md
Weight: Code 15% of aggregate score.
Agent 3: academic-proofreader (subagent_type: academic-proofreader)
Review [paper.tex] for grammar, writing quality, claims-evidence alignment,
notation consistency, and compilation. 6 check categories.
Save report to quality_reports/[file]_proofread_report.md
Weight: Paper 25% of aggregate score.
Agent 4: replication-verifier (subagent_type: replication-verifier)
Run standard verification (checks 1-4): LaTeX compilation, script execution,
file integrity, output freshness.
Save report to quality_reports/[file]_verification.md
Weight: Replication 5% of aggregate score.
If Talks/*.tex files exist, launch:
storyteller — check notation matches paper, slide count within format range. discussant — narrative flow, visual quality, content fidelity.
Talk scores are advisory, non-blocking.
Per scoring-protocol.md, compute:
Overall = Σ (weight_k × score_k) for available components
| Component | Weight | Agent |
|---|---|---|
| Literature | 10% | (skip if no lit review exists) |
| Data | 10% | (skip if no data review exists) |
| Identification | 25% | econometrics-critic |
| Code | 15% | debugger |
| Paper | 25% | academic-proofreader |
| Polish | 10% | (from academic-proofreader writing quality subscore) |
| Replication | 5% | replication-verifier |
If components are missing, renormalize weights over available components.
# Paper Excellence Report: [Title]
**Date:** [YYYY-MM-DD]
**Aggregate Score:** [XX/100] — [gate status]
## Score Breakdown
| Component | Weight | Score | Issues | Agent |
|-----------|--------|-------|--------|-------|
| Identification | 25% | XX | N | econometrics-critic |
| Code | 15% | XX | N | Debugger |
| Paper | 25% | XX | N | academic-proofreader |
| Polish | 10% | XX | N | academic-proofreader |
| Replication | 5% | XX | PASS/FAIL | replication-verifier |
## Auxiliary: Talk Scores (non-blocking)
| Talk | Score | Issues |
|------|-------|--------|
## Priority Fixes (Top 5)
1. **[CRITICAL]** [Most important — from which agent]
2. **[MAJOR]** [Second priority]
...
## Full Reports
[Links to individual report files]
Per quality-gates.md:
Sync the overleaf/ folder with the current content of paper/. Copies all LaTeX source, figures, tables, appendix, and bibliography. Puts the model section as overleaf/model.tex (top-level) for easy coauthor access. Run whenever paper/ changes and you want to push to Overleaf.
RA onboarding audit — explore all project folders and code, inspect data files directly, and produce a structured project-state document that enables a clean restart. Use at the start of a project hand-off or when picking up work from multiple prior contributors.
Full bibliography audit for an academic paper. Verifies every citation is (1) structurally correct in the .bib file, (2) actually cited in the paper, and (3) a real publication — author names, year, journal, and title verified via web search. Flags hallucinated or inaccurate references.
Design identification strategy by dispatching the causal-strategist agent (proposer) and identification-critic agent (critic). Produces strategy memo, pseudo-code, robustness plan, and falsification tests. Use when asked to "identify the effect", "design the strategy", or "write the strategy memo".
Show current context status and session health. Use to check how much context has been used, whether auto-compact is approaching, and what state will be preserved.
Show the current research pipeline status for the active project, or display the pipeline guide for a specific project type. Reads the research spec, research journal, and active plans to determine phase, scores, and next steps.