SPSS-Mac-MCP
Native macOS port of SPSS-MCP — let AI assistants drive IBM SPSS Statistics 20-31 on macOS through the Model Context Protocol. 23 analysis methods validated, triple output (OXML+SPV+HTML).
Install / Use
claude mcp add opalus12138 -- npx -y github:opalus12138/SPSS-Mac-MCPIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Education & ResearchSupported Platforms
Our assessment of SPSS-Mac-MCP
SPSS-Mac-MCP scores 77/100 on our quality scale, 407th of 444 Education & Research skills we index.
Its MCP Server is 7.0 KB long, well organised into 25 sections with 9 code examples: a thorough specification that gives an agent plenty to work with.
It has 3 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated about 5 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 90/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
SPSS-Mac-MCP compared with similar skills
All 4 of these similar skills score higher than SPSS-Mac-MCP; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| SPSS-Mac-MCP (this skill)by opalus12138 | 77 | 3 | 5mo ago | MCP Server |
| Agent-Reachby Panniantong | 100 | 106.4k | 5d ago | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 78.3k | today | CLAUDE.md |
| last30days-skillby mvanhorn | 100 | 65.9k | 1d ago | CLAUDE.md |
| CowAgentby zhayujie | 100 | 47.7k | today | CLAUDE.md |
Frequently asked questions
- How do I install SPSS-Mac-MCP?
- Run
claude mcp add opalus12138 -- npx -y github:opalus12138/SPSS-Mac-MCP. The install tabs above show the steps for each supported agent. - Which AI agents does SPSS-Mac-MCP work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is SPSS-Mac-MCP safe to use?
- It is MIT-licensed and scores 90/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is SPSS-Mac-MCP still maintained?
- The repository was last updated about 5 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
Skill content
View source on GitHubSPSS-Mac-MCP
Let AI assistants drive IBM SPSS Statistics on macOS through the Model Context Protocol. Describe your analysis in plain language — SPSS-Mac-MCP translates it into SPSS syntax, runs it on the real SPSS engine, and returns clean Markdown/SPV/HTML output.
让 AI 助手在 macOS 上直接驱动 IBM SPSS Statistics 跑统计分析。 用自然语言描述需求 —— SPSS-Mac-MCP 自动生成 SPSS 语法、调用真实 SPSS 引擎、返回 Markdown / SPV / HTML 三种格式的结果。
✨ Why SPSS-Mac-MCP
| | Windows 版 SPSS-MCP | SPSS-Mac-MCP(本项目) |
|---|---|---|
| 平台支持 | 仅 Windows | macOS + Windows + Linux |
| 安装 | install.bat | pip install -e . |
| SPSS 检测 | Windows 注册表 | macOS /Applications 自动探测 |
| 动态库处理 | PATH | DYLD_FALLBACK_LIBRARY_PATH(绕过 macOS SIP)|
| 输出格式 | OXML | OXML + SPV + HTML(三联输出)|
| 端到端验证 | 部分 | 23 项主流分析 100% 通过 |
🎯 Features
- Native macOS support:自动识别
/Applications/IBM SPSS Statistics XX/,无需手工配置 - Persistent SPSS engine:一次启动复用,避免每次冷启动 30-60 秒
- Triple-format output:同步生成机器解析的 OXML、SPSS 原生的 SPV、浏览器友好的 HTML
- 22+ SPSS methods:覆盖描述、相关、t/ANOVA/重测、线性/Logistic/多分类回归、信度/EFA、非参数、聚类、判别、生存分析
- Cross-platform compatible:同一份代码兼容 macOS、Windows、Linux
📦 Requirements
- macOS 11+(也兼容 Windows 10/11 与 Linux)
- Python 3.10+
- IBM SPSS Statistics 20–31(带 Python Essentials / Python Integration Plug-in)
- MCP 客户端:Claude Code、Claude Desktop、Cline 等
🚀 Quick Install
# 1. 用 uv 装隔离环境(推荐)
git clone https://github.com/opalus12138/SPSS-Mac-MCP.git
cd SPSS-Mac-MCP
uv venv .venv --python python3.12
uv pip install -e .
# 2. 一键写入 Claude Code 配置(自动备份原配置)
.venv/bin/spss-mac-mcp configure-claude
# 3. 重启 Claude Code
完成。Claude Code 里 /mcp 命令会列出 spss-mac-mcp。
验证安装
.venv/bin/spss-mac-mcp status
应输出:
=== SPSS MCP Capability Status ===
pyreadstat : OK v1.3.5
pandas : OK v3.0.3
SPSS batch : OK — /Applications/IBM SPSS Statistics 27/SPSS Statistics.app/Contents/MacOS/stats
💬 Usage
只需在 Claude Code 里用自然语言描述:
请对 data.sav 做描述性统计
用 SPSS 的 EFA 对 Q1-Q12 做探索性因子分析,VARIMAX 旋转,提取 2 个因子
对 firm_panel.sav 做层级回归:
Step 1: 控制变量 (Lev, ROA, Size)
Step 2: 加入 DTI
Step 3: 加入 IP
报告 R² 变化和系数表
Claude 会自动:
- 读取你的
.sav文件 - 生成 SPSS 语法
- 调用真实 SPSS 引擎执行
- 把结果整理成 Markdown 三线表(同时把
.spv留给 SPSS GUI、.html给浏览器)
🧪 Validated Analyses
本项目对 23 项主流 SPSS 分析方法做了端到端验证,全部在 macOS 上跑通:
<details> <summary>展开查看完整列表(23/23 PASS)</summary>| 类别 | 分析 | SPSS 命令 | 状态 |
|---|---|---|---|
| 描述 / 相关 / 频数 | 描述统计 | DESCRIPTIVES | ✅ |
| | 频数分布 | FREQUENCIES | ✅ |
| | 交叉表 + 卡方 | CROSSTABS | ✅ |
| | Pearson 相关 | CORRELATIONS | ✅ |
| | Spearman 等级相关 | NONPAR CORR | ✅ |
| 均值比较 | 独立样本 t | T-TEST GROUPS | ✅ |
| | 配对 t | T-TEST PAIRS | ✅ |
| | 单因素 ANOVA | ONEWAY | ✅ |
| | 双因素 ANOVA | UNIANOVA | ✅ |
| | 重复测量 ANOVA | GLM w/ WSFACTOR | ✅ |
| 回归 | 线性回归 | REGRESSION | ✅ |
| | 层级回归 | REGRESSION 多步 | ✅ |
| | 二元 Logistic | LOGISTIC REGRESSION | ✅ |
| | 多分类 Logistic | NOMREG | ✅ |
| 量表方法 | Cronbach's α | RELIABILITY | ✅ |
| | EFA 因子分析 | FACTOR w/ VARIMAX | ✅ |
| 非参数检验 | Mann-Whitney U | NPAR TESTS /M-W | ✅ |
| | Kruskal-Wallis | NPAR TESTS /K-W | ✅ |
| | Wilcoxon 符号秩 | NPAR TESTS /WILCOXON | ✅ |
| 多元方法 | K-均值聚类 | QUICK CLUSTER | ✅ |
| | 层次聚类 | CLUSTER | ✅ |
| | 判别分析 | DISCRIMINANT | ✅ |
| 生存分析 | Kaplan-Meier | KM | ✅ |
完整演示见 examples/full_validation/。
🔧 Technical Highlights
macOS 适配的几个关键决策
DYLD_FALLBACK_LIBRARY_PATH而非DYLD_LIBRARY_PATH:macOS SIP 在 Catalina+ 之后会剥掉子进程的DYLD_LIBRARY_PATH,但DYLD_FALLBACK_LIBRARY_PATH不受影响。- 绕过 IBM 硬编码的
/Applications/Python3/lib/:SPSS 自带的python3.8二进制把libpython3.8.dylib路径硬编码到/Applications/Python3/(一个非默认位置)。通过DYLD_FALLBACK_LIBRARY_PATH=$SPSS_HOME/Resources/Python3/lib让它能找到真实的库。 - OMS 直接输出 SPV:XD 模式(Python 启动的外部 SPSS 引擎)下
OUTPUT SAVE/OUTPUT EXPORT不可用,但OMS /DESTINATION FORMAT=SPV可以。
详细技术报告见 docs/macos-port.md。
📊 Output Formats
每次分析自动生成三种格式:
| 格式 | 用途 | 推荐场景 |
|---|---|---|
| OXML (.xml) | 机器解析 | 写脚本提取系数/p 值/R² |
| SPV (.spv) | SPSS Viewer 原生 | 在 SPSS GUI 打开、改透视表样式、复制到 Word |
| HTML (.html) | 浏览器 | 直接看、分享给同事、贴到邮件 |
🧠 PROCESS Macro Integration
集成 Andrew Hayes 的 PROCESS v4.1,覆盖 6 大常用模型(已端到端验证,见 examples/process_models/)。
用法
# 通过 MCP 工具
spss_run_process(
file_path="data.sav",
y="TQ", x="DTI", m=["IP"], # 注意:m 必须是列表
model=4, bootstrap=5000, seed=20260525,
)
或者用自然语言:
"用 PROCESS Model 4 跑 firm_panel.sav 的 DTI → IP → TQ 中介,
bootstrap 5000 次,标准化系数"
⚠️ FAQ: 常见错误
错误 1:MATRIX Error #12302 或 PROCESS 解析失败
症状:直接用 spss_run_syntax 自己写 PROCESS 调用语法时报错。
根因:用了错误的 SPSS 命令 INCLUDE FILE='...'.
正确做法:
INSERT FILE='/path/to/process.sps'. ← ✅ 用 INSERT
PROCESS y=Y /x=X /m=M /model=4 /boot=5000.
# 而不是:
INCLUDE FILE='/path/to/process.sps'. ← ❌ 不能用 INCLUDE
为什么?INCLUDE 是 SPSS 老命令,对长行有 80 字符截断,会破坏 PROCESS 内部的 MATRIX ... END MATRIX 块。Andrew Hayes 官方文档明确要求用 INSERT。
如果你用本项目的 spss_run_process 工具,已经自动用 INSERT,不用操心这个。
错误 2:调用 spss_run_process 报 "Missing required argument"
症状:参数明明传了,但报缺 file_path / y / x。
根因:早期版本(≤ v0.1.x)的 schema 用了过于宽松的类型(object、空 list/dict),客户端 XML 序列化器会把整个 dict 渲染成空。
修复:v0.2.0+ 已经把 m 改成明确的 List[str],covariates 为 List[str],extra_options 为 Dict[str, str]。升级即可。
错误 3:输出超大(900KB+)超过 LLM 上下文
症状:PROCESS 跑完后输出几百 KB 的源码 echo,把 LLM 上下文塞爆。
根因:PROCESS 用 MATRIX 引擎,源码会被完整 echo。SET PRINTBACK=OFF 在 XD 模式下无效。
修复:v0.2.0+ 内置 extract_process_results() 后处理,自动定位最后一次"PROCESS Procedure"banner,把输出从 ~937 KB 压到 ~3 KB(99.7% 压缩)。
🙏 Acknowledgments
This project is derived from SPSS-MCP by Exekiel179 (MIT licensed).
The original project pioneered the persistent SPSS engine + OMS capture architecture, which made the macOS port feasible. SPSS-Mac-MCP focuses specifically on:
- Native macOS adaptation (path detection, DYLD handling)
- Triple-format output (OXML + SPV + HTML)
- Comprehensive validation of 23 analysis methods
- Cross-platform unification (one codebase, three OSes)
IBM® and SPSS® are trademarks of International Business Machines Corporation. This project is not affiliated with, endorsed by, or sponsored by IBM.
📜 License
MIT — see LICENSE.
Acknowledgment notice in NOTICE.
🛠️ Contributing
Issues 与 PR 欢迎。重点方向:
- 接入更多 SPSS 命令(PROCESS、SEM AMOS 桥接)
- Windows 同步验证(保持跨平台兼容)
- 性能优化(多分析并行)
Author: opalus12138
Related Skills
Agent-Reach
106.4kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
78.3kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
last30days-skill
65.9kAI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary
CowAgent
47.7kOpen-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
