grahamrowe82/mcp-quality-index
收藏Hugging Face2026-03-27 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/grahamrowe82/mcp-quality-index
下载链接
链接失效反馈官方服务:
资源简介:
---
license: cc-by-4.0
task_categories:
- text-classification
- feature-extraction
tags:
- mcp
- model-context-protocol
- open-source
- quality-scores
- developer-tools
- ai-ecosystem
pretty_name: MCP Quality Index
size_categories:
- 10K<n<100K
source_datasets:
- original
---
# MCP Quality Index
Daily-updated quality scores for 12,000+ MCP (Model Context Protocol) server repositories.
## Dataset description
Every MCP registry today is a flat catalog. None of them tell you whether a server is maintained, adopted, or safe to depend on. This dataset scores every MCP-domain repository on GitHub across four dimensions: maintenance, adoption, maturity, and community.
## Files
| File | Records | Description |
|---|---|---|
| `mcp-scores.json` | 12,653 | Quality scores with component breakdown and risk flags |
| `mcp-repos.json` | 12,512 | All active MCP repos with GitHub + package metrics |
| `projects.json` | 441 | Tracked AI projects with traction scores and velocity |
## Scoring model
Each repo receives a composite quality score (0-100) from four equally-weighted components:
- **Maintenance** (0-25): commit activity + push recency
- **Adoption** (0-25): stars + downloads + reverse dependents
- **Maturity** (0-25): license + published package + repo age
- **Community** (0-25): forks + fork-to-star ratio
Scores map to tiers: **Verified** (70+), **Established** (50-69), **Emerging** (30-49), **Experimental** (<30).
Full methodology: [METHODOLOGY.md](https://github.com/grahamrowe82/mcp-quality-index/blob/main/METHODOLOGY.md)
## Source
Data is produced by [PT-Edge](https://github.com/grahamrowe82/pt-edge), which tracks 166,000+ AI repositories across GitHub, PyPI, npm, Docker Hub, HuggingFace, and Hacker News.
## Update frequency
Daily at approximately 06:00 UTC.
## Citation
```bibtex
@misc{mcp-quality-index-2026,
title={MCP Quality Index},
author={PT-Edge},
year={2026},
url={https://github.com/grahamrowe82/mcp-quality-index}
}
```
许可证:CC BY 4.0
任务类别:
- 文本分类
- 特征提取
标签:
- mcp
- 模型上下文协议(Model Context Protocol)
- 开源
- 质量评分
- 开发者工具
- AI生态系统
展示名称:MCP质量指数
规模区间:10000 < 条目数 < 100000
源数据集:
- 原创数据集
---
# MCP质量指数
本数据集提供12000余个模型上下文协议(Model Context Protocol,简称MCP)服务器仓库的每日更新质量评分。
## 数据集描述
当前所有MCP注册平台均为扁平化目录,均无法告知用户某一服务器是否处于维护状态、是否被广泛采用,或可供安全依赖。本数据集针对GitHub上所有属于MCP领域的仓库,从维护性、采用度、成熟度与社区活跃度四个维度进行评分。
## 文件列表
| 文件名称 | 条目数 | 描述 |
|---|---|---|
| `mcp-scores.json` | 12653 | 包含分项评分与风险标记的质量评分文件 |
| `mcp-repos.json` | 12512 | 收录所有活跃MCP仓库,附带GitHub与包管理指标 |
| `projects.json` | 441 | 收录受追踪的AI项目,附带吸引力评分与发展速度指标 |
## 评分模型
每个仓库将获得由四个权重均等的分项构成的综合质量评分(分值区间0-100):
- **维护性**(0-25分):提交活跃度与推送时效性
- **采用度**(0-25分):星标数、下载量与反向依赖项数量
- **成熟度**(0-25分):许可证类型、已发布包状态与仓库存续时长
- **社区活跃度**(0-25分):复刻数量与复刻-星标比率
评分将对应以下等级划分:**已验证**(70分及以上)、**成熟稳定**(50-69分)、**崭露头角**(30-49分)与**实验性**(低于30分)。
完整方法论文档详见:[METHODOLOGY.md](https://github.com/grahamrowe82/mcp-quality-index/blob/main/METHODOLOGY.md)
## 数据来源
本数据集由[PT-Edge](https://github.com/grahamrowe82/pt-edge)项目生成,该项目追踪了GitHub、PyPI、npm、Docker Hub、HuggingFace与Hacker News平台上共计166000余个AI仓库。
## 更新频率
每日约协调世界时(UTC)06:00更新。
## 引用格式
bibtex
@misc{mcp-quality-index-2026,
title={MCP Quality Index},
author={PT-Edge},
year={2026},
url={https://github.com/grahamrowe82/mcp-quality-index}
}
提供机构:
grahamrowe82



