Skip to content
Conduit

skill

paper-quality-filter

A bilingual Agent Skill and local dataset that helps AI agents rank academic papers using JCR, impact factor, CAS zones, CCF ranks, EI indexing, and Chinese core journal evidence.

yilaai/paper-quality-filtercommunitylistedunreviewed
Use skill

Set up

Skills are installed by placing their folder where your agent loads skills.

Download ZIP from GitHub Open the repository

The download comes from github.com/yilaai/paper-quality-filter itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 22, 2026

Paper Quality Filter

Make AI agents understand academic venue quality.
让 AI Agent 读懂期刊分区、会议等级、EI 和中文核心。

中文说明

AI can read papers, but it does not always understand publication venue quality.
This Skill gives AI agents the evidence layer they need.

Paper Quality Filter is a bilingual Agent Skill and local reference dataset for screening and ranking academic papers by publication venue evidence.

When researchers ask ChatGPT, Claude, Codex, or other AI agents to screen papers, write literature reviews, or inspect references, the model may understand the topic but may not reliably know whether a journal is JCR Q1, whether a conference is CCF A, whether a source is EI-indexed, or whether a Chinese journal is 北大核心, CSSCI, AMI核心, or 科技核心.

This project gives AI agents structured local evidence, so they do not have to guess from model memory.

What Is Included?

DatasetRecordsWhat it helps AI understand
Journal scores22,496JCR quartile, impact factor, CAS zone
CCF conferences971CCF A/B/C conference ranking
EI sources5,269EI-indexed sources
Chinese core journal tags4,594北大核心, CSSCI, AMI核心, 科技核心

Use Cases

Use this Skill when you want an AI agent to:

  • rank papers before writing an AI literature review;
  • screen references from Google Scholar, PubMed, Semantic Scholar, Web of Science, CNKI, Zotero, EndNote, or manual lists;
  • explain why a paper should be read first using venue-level evidence;
  • identify JCR Q1/Q2 journals, high-impact journals, and CAS Zone 1/2 journals;
  • recognize CCF A/B/C conferences in computer science;
  • identify EI-indexed sources;
  • screen Chinese papers using 北大核心, CSSCI, AMI核心, and 科技核心 tags;
  • add a venue-quality evidence layer to a research agent, academic writing AI, RAG workflow, or literature review agent.

Input Formats

The ranking script accepts:

  • .json
  • .csv
  • .md
  • .txt
  • .bib / BibTeX
  • .ris / RIS-like exports

Useful fields include title, venue, journal, conference, source, container-title, publication, booktitle, year, and doi.

Quick Start

python scripts/rank_papers.py examples/papers.csv --lang en --format markdown
python scripts/rank_papers.py examples/chinese-papers.md --lang zh --format markdown
python scripts/rank_papers.py examples/bibtex-sample.bib --lang en --top 20 --format json --output ranked.json

Recommendation Tiers

TierChineseTypical Evidence
Strongly Recommended强烈推荐CCF A, CAS Zone 1, JCR Q1 with IF >= 10
Worth Keeping值得保留CCF B, JCR Q1, CAS Zone 2, IF >= 5
Supplementary可作为补充CCF C, JCR Q2, EI, 北大核心, CSSCI, AMI核心, 科技核心
Insufficient Venue Evidence质量依据不足Missing venue, unmatched source, preprint-only source

“Insufficient Venue Evidence” does not mean the paper is low quality. It only means the current local reference data did not confirm the publication venue.

Output Example

Processed 42 papers:

- Strongly Recommended: 8
- Worth Keeping: 15
- Supplementary: 10
- Insufficient Venue Evidence: 9

Coverage:
- Papers with matched venue evidence: 33
- Papers without matched venue evidence: 9
- Match rate: 78.6%

| Title | Venue | Year | Recommendation | Evidence | Confidence |
| --- | --- | --- | --- | --- | --- |
| Example Paper | Nature Machine Intelligence | 2024 | Strongly Recommended | JCR Q1; IF 23.9; CAS Zone 1 | high |

Agent Skill

This repository includes a SKILL.md file for Codex-style Agent Skill usage. The Skill tells an agent when to use the dataset, how to extract venue metadata, how to run the script, and how to explain ranking results without overclaiming.

Example prompt:

Help me screen this paper list and rank the papers by venue-level quality evidence.

中文示例：

帮我从这批论文中筛选高质量论文，并说明每篇论文的推荐依据。

Limitations

This project does not judge the full scientific contribution of a paper.

A paper from a top journal can still be irrelevant to your topic. A paper from an unmatched venue can still be valuable. This Skill only provides venue-level screening signals. Final academic judgment should still consider topical relevance, methodology, data quality, citation context, publication year, and whether the paper actually supports your claim.

Documentation

Built by yila.ai

This project is maintained by yila.ai, an academic agent workspace for researchers and students.

yila.ai helps with paper discovery, literature review writing, scientific figure generation, academic writing support, and other core research workflows.

License

Code is released under the MIT License. Please review repository notes and data-source constraints before redistributing or using the dataset commercially.

↑↓ navigate openesc close