MCPSolver-Eval

MCP运维云端

提供方 JunseokTheLee

运维云端 stdio · 本地进程 API Key 社区 / 第三方 ClaudeGemini CLIZed
查看接入方式

The MCP Solver is a server that exposes constraint solving, SAT, SMT, and ASP capabilities to Large Language Models via the Model Context Protocol, suitable for complex problem scenarios requiring symbolic reasoning and constraint resolution.

一句话选型结论

按资料字段整理

MCPSolver-Eval 属于运维云端类 MCP 服务器,stdio · 本地进程,API Key,可直接接入 Claude、Gemini CLI、Zed,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类运维云端 传输方式stdio · 本地进程 鉴权API Key 部署方式本地部署
GitHub 星标 0 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

运维云端
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

Local Deployment Constraint Solving Symbolic Reasoning Python SAT Solver SMT Solver ASP Solver Model Context Protocol

支持客户端

适用场景

Local DeploymentConstraint SolvingSymbolic Reasoning

安装与配置

stdio · 本地进程 API Key
客户端在本机拉起子进程,数据不出本机
git clone https://github.com/szeider/mcp-solver.git cd mcp-solver uv venv source .venv/bin/activate uv pip install -e ".[all]" # Install all solvers

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 python,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
云资源查询

查实例、存储、账单等云上资源状态。

部署与发布

把部署、重启、回滚这类操作接进对话。

监控与告警

拉取监控指标,异常时让模型帮你判断。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude、Gemini CLI、Zed,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 使用前要准备好API Key对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在运维云端场景里让模型直接调用外部能力的人。
  • 已经在用 Claude、Gemini CLI 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
MCPSolver-Eval当前 未标注 stdio · 本地进程 本地部署 有 2026-06-07
lobehub 未标注 stdio · 本地进程 本地部署 有 2026-04-11
TrendRadar 未标注 stdio · 本地进程 本地部署 有 2026-04-13
Ui Tars Desktop 8a825770 未标注 stdio · 本地进程 本地部署 有 2026-05-17
ruflo 未标注 stdio · 本地进程 本地部署 有 2026-04-13
FastGPT 未标注 stdio · 本地进程 本地部署 有 2026-04-11

服务器档案

运维云端
字段来自项目仓库与公开索引,中文说明由本站标注
提供方JunseokTheLee上游标注的维护方 JunseokTheLee
分类运维云端本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式API Key远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · python本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-06-07上游最后核验时间

详细介绍

9.9 KB
来自项目仓库与上游目录的原文,未做删改

MCP Solver

A Model Context Protocol (MCP) server that exposes constraint solving, SAT, SMT, and ASP capabilities to Large Language Models.


Overview

The MCP Solver integrates constraint solving, SAT, SMT, and ASP with LLMs through the Model Context Protocol, enabling AI models to interactively create, edit, and solve:

  • Constraint models in MiniZinc
  • SAT models in PySAT
  • MaxSAT optimization problems in PySAT
  • SMT formulas in Z3 Python
  • Answer Set Programs in Clingo

For a detailed description of the MCP Solver's system architecture and theoretical foundations, see the accompanying research paper: Stefan Szeider, "Bridging Language Models and Symbolic Solvers via the Model Context Protocol", SAT 2025.

Available Tools

In the following, item refers to some part of the (MiniZinc/PySAT/Z3/ASP) code, and model to the encoding.

Tool Name Description
clear_model Remove all items from the model
add_item Add new item at a specific index
delete_item Delete item at index
replace_item Replace item at index
get_model Get current model content with numbered items
solve_model Solve the model (with timeout parameter)

System Requirements

  • Python and project manager uv
  • Python 3.11+
  • Mode-specific requirements: MiniZinc, PySAT, Python Z3 (required packages are installed via pip)
  • Operating systems: macOS, Windows, Linux (with appropriate adaptations)

Installation

MCP Solver requires Python 3.11+, the uv package manager, and solver-specific dependencies (MiniZinc, Z3, or PySAT).

For detailed installation instructions for Windows, macOS, and Linux, see INSTALL.md.

Quick start:```bash
git clone https://github.com/szeider/mcp-solver.git
cd mcp-solver
uv venv
source .venv/bin/activate
uv pip install -e ".[all]" # Install all solvers


------

## Available Modes / Solving Backends

The MCP Solver provides five distinct operational modes, each integrating with a different constraint solving backend. Each mode requires specific dependencies and offers unique capabilities for addressing different classes of problems.

### MiniZinc Mode

MiniZinc mode provides integration with the MiniZinc constraint modeling language with the following features:

- Rich constraint expression with global constraints
- Integration with the Chuffed constraint solver
- Optimization capabilities
- Access to solution values via `get_solution`

**Dependencies**: Requires the `minizinc` package (`uv pip install -e ".[mzn]"`)

**Configuration**: To run in MiniZinc mode, use:```
mcp-solver-mzn

PySAT Mode

PySAT mode allows interaction with the Python SAT solving toolkit with the following features:

  • Propositional constraint modeling using CNF (Conjunctive Normal Form)
  • Access to various SAT solvers (Glucose3, Glucose4, Lingeling, etc.)
  • Cardinality constraints (at_most_k, at_least_k, exactly_k)
  • Support for boolean constraint solving

Dependencies: Requires the python-sat package (uv pip install -e ".[pysat]")

Configuration: To run in PySAT mode, use:```
mcp-solver-pysat


### MaxSAT Mode

MaxSAT mode provides specialized support for optimization problems with PySAT, featuring:

- Weighted Conjunctive Normal Form (WCNF) support
- Integration with the RC2 MaxSAT solver
- Optimization capabilities with objective tracking
- Support for both hard and soft constraints

**Dependencies**: Requires the `python-sat` package (`uv pip install -e ".[pysat]"`)

**Configuration**: To run in MaxSAT mode, use:```
mcp-solver-maxsat

Z3 Mode

Z3 mode provides access to Z3 SMT (Satisfiability Modulo Theories) solving capabilities with the following features:

  • Rich type system: booleans, integers, reals, bitvectors, arrays
  • Constraint solving with quantifiers
  • Optimization capabilities
  • Template library for common modeling patterns

Dependencies: Requires the z3-solver package (uv pip install -e ".[z3]")

Configuration: To run in Z3 mode, use:```
mcp-solver-z3


### ASP Mode

ASP (Answer Set Programming) mode provides integration with ASP solvers (e.g., Clingo) for declarative problem solving with logic programs. Features include:

- Expressive logic programming for combinatorial and knowledge representation problems
- Support for constraints, choice rules, aggregates, and optimization statements
- Integration with the Clingo solver for efficient answer set computation
- Access to answer sets and model inspection

**Dependencies**: Requires the `clingo` package (`uv pip install -e ".[asp]"` or included in `[all]`)

**Configuration**: To run in ASP mode, use:```
mcp-solver-asp

MCP Test Client

Installation```bash

Install client dependencies

uv pip install -e ".[client]"

Verify client installation and configuration

uv run test-setup-client


The client requires an **API key** from an LLM provider. For Anthropic (the default LLM is Claude Sonnet 3.7), set the `ANTHROPIC_API_KEY` environment variable. This can be set in your environment or included in a `.env` file in the project root.

The client also supports other LLM providers including OpenAI, Google (Gemini), OpenRouter and even local models via LM Studio. You can specify which model to use with the `--mc` command line flag. See INSTALL.md for details on model code formats.

### Usage

The client provides specialized commands for each solver backend:```bash
# MiniZinc mode
uv run run-test mzn --problem <path/to/problem.md>

# PySAT mode
uv run run-test pysat --problem <path/to/problem.md>

# MaxSAT mode
uv run run-test maxsat --problem <path/to/problem.md>

# Z3 mode
uv run run-test z3 --problem <path/to/problem.md>

# ASP mode
uv run run-test asp --problem <path/to/problem.md>

Examples

Example 1: Casting Problem (MiniZinc)

This is based on an example by George Johnson (NYT, July 13, 1999).

User

> Suppose that a theatrical director feels obligated to cast either his ingenue, Actress Alvarez, or his nephew, Actor Cohen, in a production. But Miss Alvarez won't be in a play with Mr. Cohen (her former lover), and she demands that the cast include her new flame, Actor Davenport. The producer, with her own favors to repay, insists that Actor Branislavsky have a part. But Mr. Branislavsky won't be in any play with Miss Alvarez or Mr. Davenport. Can the director cast the play?

System: clear_model

>> Model cleared >

System: add_item (multiple)

>minizinc > var bool: alvarez; > var bool: cohen; > var bool: branislavsky; > var bool: davenport; > constraint alvarez \/ cohen; % Must cast either Alvarez or Cohen > constraint not (alvarez /\ cohen); % Alvarez won't work with Cohen > constraint alvarez -> davenport; % If Alvarez is cast, Davenport must be cast > constraint branislavsky; % Branislavsky must be cast > constraint not (branislavsky /\ alvarez); % Branislavsky won't work with Alvarez > constraint not (branislavsky /\ davenport); % Branislavsky won't work with Davenport > solve satisfy; >

System: solve_model

>> {'status': 'SAT', 'solution': Solution(alvarez=False, cohen=True, branislavsky=True, davenport=False, _checker=''), 'solve_time': 0.0} >

The solution (cast Cohen and Branislavsky, but not Alvarez or Davenport) satisfies all constraints.

Example 2: N-Queens Problem (MiniZinc)

User

> Check whether you can place n Queens on an nxn chessboard. Try n=10,20,30,40 and compare the solving times

System: clear_model

>> Model cleared >

System: add_item (multiple)

>minizinc > int: n = 10; > array[1..n] of var 1..n: queens; > constraint alldifferent(queens); > constraint alldifferent([queens[i] + i | i in 1..n]); > constraint alldifferent([queens[i] - i | i in 1..n]); > solve satisfy; >

System: solve_model (with n=10, 20, 30, and 40 by changing the value of n)

Solve times comparison: n=10: 0.001s n=20: 0.005s n=30: 0.040s n=40: 0.043s

Solutions exist for all tested board sizes. Time complexity increases non-linearly but remains manageable up to n=40.

Example 3: Traveling Salesperson Problem (MiniZinc)

User

> A saleswoman based in Vienna needs to plan her upcoming tour through Austria, visiting each province capital once. Help find the shortest route. Distances in km: 1=Vienna, 2=St. Pölten, 3=Eisenstadt, 4=Linz, 5=Graz, 6=Klagenfurt, 7=Salzburg, 8=Innsbruck, 9=Bregenz
>
> [Distance matrix table provided in user message]

System: add_item (multiple)

>minizinc > include "globals.mzn"; > int: n = 9; > array[1..n,1..n] of int: dist = [|0,65,60,184,195,319,299,478,631|65,0,125,119,130,254,234,413,566|60,125,0,184,157,281,261,440,593|184,119,184,0,208,252,136,315,468|195,130,157,208,0,136,280,459,629|319,254,281,252,136,0,217,391,566|299,234,261,136,280,217,0,188,343|478,413,440,315,459,391,188,0,157|631,566,593,468,629,566,343,157,0|]; > array[1..n] of var 1..n: succ; > var int: total_dist = sum(i in 1..n)(dist[i,succ[i]]); > constraint circuit(succ); > solve minimize total_dist; >

System: solve_model

>> {'status': 'SAT', 'solution': Solution(objective=1564, succ=[2, 4, 1, 7, 3, 5, 9, 6, 8], _checker=''), 'solve_time': 0.002} >

Optimal route: Vienna → St. Pölten → Linz → Salzburg → Bregenz → Innsbruck → Klagenfurt → Graz → Eisenstadt → Vienna with a total distance of 1,564 km.

Feedback

You can provide feedback to the author via this form.

Disclaimer

This MCP Solver is in its prototype stage and should be used with caution. Users are encouraged to experiment, but any use in critical environments is at their own risk.


License

This project is licensed under the MIT License - see the LICENSE file for details.


常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
MCPSolver-Eval 怎么安装?

在客户端按下面的安装命令接入:git clone https://github.com/szeider/mcp-solver.git cd mcp-solver uv venv source .venv/bin/activate uv pip install -e ".[all]" # Install all solvers。装完在客户端配置里引用同一个包或端点。

MCPSolver-Eval 怎么在 Claude、Gemini CLI、Zed 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

MCPSolver-Eval 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

MCPSolver-Eval 需要 API Key 吗?

资料标注需要API Key认证,使用前准备好对应凭据,并按最小权限配置。

MCPSolver-Eval 支持哪些客户端?

资料标注支持 Claude、Gemini CLI、Zed;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

MCPSolver-Eval 能提供哪些能力?

资料显示能力标签包括 Local Deployment、Constraint Solving、Symbolic Reasoning、Python、SAT Solver、SMT Solver。具体工具清单以官方文档为准。

MCPSolver-Eval 需要什么运行环境?

运行时为 python;属于本地进程,需要在本机装好运行环境。

MCPSolver-Eval 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-06-07。