mcpField
中
中国社保与薪税计算器 (China Shebao & Payroll Tax MCP)

中国社保与薪税计算器 (China Shebao & Payroll Tax MCP)

Remote
shebao.orgFinance0Updated 6d ago

About 中国社保与薪税计算器 (China Shebao & Payroll Tax MCP)

China Shebao & Payroll Tax MCP provides tools for checking Chinese social insurance and housing fund policies and estimating payroll taxes, bonuses, and pensions. It is intended for individuals and payroll planners, with nationwide policy data and stateless calculations.

Key features

  1. 01Query social security policies
  2. 02Check city fund contribution limits
  3. 03Estimate take-home pay and deductions
  4. 04Calculate income and bonus tax
  5. 05Estimate retirement pension

Use cases

  1. 01Check local contribution limits and policy sources
  2. 02Estimate monthly take-home pay
  3. 03Compare annual bonus tax methods
  4. 04Project monthly retirement pension

MCP support

It exposes six policy lookup and calculation tools through a remote Streamable HTTP endpoint. A local stdio bridge is also available for clients that only support local processes.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://www.shebao.org/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

中国社保与薪税计算器 MCP Server (China Shebao & Payroll Tax MCP)

MCP Registry License: MIT Website OpenAPI 3.1

社保薪税通 (shebao.org) 是中国领先的纯前端无状态薪税、五险一金与社保政策计算平台,提供全国 31 个省市自治区官方政策、392 个地级及以上城市五险一金标准与个人所得税智能测算引擎。

本仓库是社保薪税通的官方 Model Context Protocol (MCP) 开源集成:

  1. 远程 Streamable HTTP 服务端点(推荐,零安装免鉴权):https://www.shebao.org/mcp
  2. 轻量本地 stdio 桥接工具:npx -y github:ddter/china-shebao-mcp,为仅支持本地进程的客户端提供极速转发。

快速接入 (Quick Start)

方案一:远程 Streamable HTTP(推荐,免安装)

直接将支持 Streamable HTTP (协议版本 2025-06-18 / 2025-03-26 / 2026-07-28) 的客户端指向端点:

https://www.shebao.org/mcp
1. Claude Code
claude mcp add --transport http china-shebao https://www.shebao.org/mcp
2. Cursor / VS Code / Windsurf (.cursor/mcp.json 或 mcp.json)
{
  "mcpServers": {
    "china-shebao": {
      "url": "https://www.shebao.org/mcp"
    }
  }
}
3. OpenAI Codex CLI (~/.codex/config.toml)
[mcp_servers.china-shebao]
url = "https://www.shebao.org/mcp"
4. ChatGPT Custom GPT Actions

在 GPT 编辑器中点击 Configure → Create new action,在 Import from URL 填入:

https://www.shebao.org/openapi.json

即可一键导入全量薪税社保测算接口。


方案二:本地 stdio 桥接脚手架

针对部分只支持 stdio 标准输入输出的客户端(例如 Claude Desktop),在 claude_desktop_config.json 中配置:

{
  "mcpServers": {
    "china-shebao": {
      "command": "npx",
      "args": ["-y", "github:ddter/china-shebao-mcp"]
    }
  }
}

提供的 6 大核心测算工具 (Tools)

工具名称功能说明核心输入参数
query_social_security_policy查询 31 省市 2026 官方社保基数上限、下限、社平工资及发文依据province(省份名)
get_city_social_fund_limits查询全国 392 城五险一金基数上下限、默认缴存比例及官方核验状态city(城市名或拼音 slug)
calculate_salary_take_home测算月薪/年薪税后到手工资、五险一金明细、累计预扣个税及 12 个月现金流分布city, salary, fundRate, specialDeduction
calculate_income_tax工资薪金累计预扣、劳务报酬按次预扣、年度综合所得汇算清缴type, amount, month, annualLabor
compare_annual_bonus_tax对比年终奖单独计税与合并计税两种方案的个税差异与到手金额,提供最优筹划建议annualSalary, bonus
calculate_retirement_pension企业职工基本养老保险退休金测算(月基础养老金 + 个人账户养老金)pensionBase, years, accountBalance, retireAge

示例对话场景 (Example Prompts)

  • 场景 1(查政策):“帮我查一下 2026 年北京最新的社保缴费基数上下限和官方发文依据。”
  • 场景 2(算到手):“我在上海月薪 15,000 元,公积金缴 7%,每月专项附加扣除 1,500 元,税后能拿到手多少?五险一金公司和个人各交多少?”
  • 场景 3(年终奖筹划):“全年工资 18 万,年底发 5 万元年终奖,应该单独计税还是并入综合所得更划算?能帮我算出税额差异吗?”
  • 场景 4(养老金退休试算):“计发基数 8,000 元,累计工龄 25 年,个人账户累计 12 万元,60 岁退休每月能领多少养老金?”

数据准确性与合规说明

  1. 官方核验覆盖:全国 392 个城市的社保缴费基数上下限均已与各省人社厅、医保局 2026 年度最新政策文件逐条核验,数据来源清晰可溯源至政府门户网站(*.gov.cn)。
  2. 公积金核验透明度:重点城市公积金基数与默认缴存比例已核验官方通知,未核验城市统一采用推荐比例(5%)明确试算边界。
  3. 隐私安全:全站无状态计算,不存储、不上传任何个人姓名、证件号码或薪酬明细。
  4. 免责声明:测算结果仅供薪酬筹划与参考,实际扣缴与待遇发放以当地社保局、公积金中心及税务机关核定为准。

开源协议

本项目遵循 MIT License 开源协议。 官方网站:https://www.shebao.org/

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.