nsfc-reviewers

Reviews NSFC grant proposals with parallel multi-expert panels and outputs prioritized revision reports.

2.7k|251|Updated Feb 12, 2024
One-click install
npx skills add https://github.com/huangwb8/ChineseResearchLaTeX --skill nsfc-reviewers
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nsfc-reviewers
Source: https://github.com/huangwb8/ChineseResearchLaTeX/tree/main/skills/nsfc-reviewers
Command: npx skills add https://github.com/huangwb8/ChineseResearchLaTeX --skill nsfc-reviewers

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Researchers preparing NSFC (National Natural Science Foundation of China) proposals lack access to realistic pre-submission peer review, making it hard to identify fatal flaws, gauge funding risk, and prioritize revisions before the actual deadline.

Core Features & Use Cases

  • Multi-Expert Parallel Review: Simulates panels of 7 expert personas (innovation, methodology, foundation, critical, constructive, significance, clarity) across 1-5 independent review groups, with cross-panel consensus aggregation.
  • Binary Stage Verdicts: Forces explicit pass/fail judgments for both letter review (函评) and panel review (会评) based on the current draft, with confidence levels and key turnaround actions.
  • Funding-Constraint Awareness: Distinguishes genuine design errors from compromises forced by budget limits (e.g., 30-40w youth fund vs 50-60w general program), and shows what an unconstrained design would look like.
  • Use Case: Before submitting a youth fund application, run a 3-panel review on your LaTeX proposal directory to receive a P0/P1/P2-prioritized issue list with evidence anchors and a minimal revision sequence.

Quick Start

Ask the AI to review your NSFC proposal directory with three expert panels and save the aggregated report into the proposal folder.

Frequently Asked Questions about nsfc-reviewers

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review an NSFC proposal before submission?

Point the skill at your proposal directory containing .tex files and request a review. It runs 7 expert personas per panel, aggregates findings across panels, and outputs a report with P0/P1/P2 issues, evidence anchors, and pass/fail verdicts for letter and panel review stages.

How many review panels should I use for grant proposal review?

Use 1 panel (7 experts) for early drafts, 2-3 panels for pre-submission checks, and up to 5 panels for critical proposals. Cross-panel consensus requires at least 60% of panels to flag the same issue, so more panels increase confidence but cost more tokens.

Does the NSFC review skill modify or compile my LaTeX files?

No. The skill is strictly read-only: it reads .tex files, analyzes content, and writes a separate review report. It never compiles LaTeX, edits proposal source files, or sends content externally.

What happens if parallel-vibe is not installed?

The skill automatically degrades to single-panel mode when parallel-vibe is unavailable, disabled, or panel_count is set to 1. You still get all 7 expert personas reviewing independently, with aggregation performed within that single panel.

Why does the review report mention funding amount constraints?

The skill distinguishes real design errors from compromises forced by budget limits, such as youth fund (30-40w) versus general program (50-60w) scopes. Constrained weaknesses are labeled with their root cause plus a reference design showing what an unconstrained version would include.