Token导航 LogoToken导航TokenDH.com
中转站排行榜
Nanohub MCP logo
AI代理stdio官方级别未说明来源级核验

Nanohub MCP

MCP Server

一个零依赖的Python库，用于创建与nanoHUB/HubZero工具基础设施集成的Model Context Protocol (MCP)服务器。

工具数

2

提示词数

0

GitHub Stars

0

资源数

0
服务器开发Python工具集成

安装说明

本站只整理中文说明和来源信息，不托管安装包，也不代用户安装。

GitHub

作者 / 组织

denphi

提供方

denphi

最后核验

2026/5/17 20:20

快速接入

先看主来源和安装命令，再打开仓库或文档；下面只保留这个条目的关键接入事实。

命令预览

pip install nanohub-mcp

详细介绍

nanohub mcp

用于创建的零依赖Python库 模型上下文协议（MCP） 与nanoHUB/HubZero工具基础架构集成的服务器。

特征：

  • 零外部依赖（仅限stdlib）
  • Python 3.7+兼容
  • SSE和流式HTTP传输
  • OpenAPI模式自动生成
  • 直接REST风格的工具调用
  • nanoHUB代理集成开箱即用
  • 用于日志记录和进度报告的上下文注入
  • 用于长时间运行作业的异步工具（无代理超时）

安装

pip install nanohub-mcp

快速开始

创建一个名为的文件 start_mcp.py:

from nanohubmcp import MCPServer

server = MCPServer("my-calculator", version="1.0.0")

@server.tool()
def add(a, b):
    # type: (float, float) -> float
    """Add two numbers together."""
    return float(a) + float(b)

@server.tool()
def multiply(a, b):
    # type: (float, float) -> float
    """Multiply two numbers."""
    return float(a) * float(b)

@server.resource("config://settings")
def get_settings():
    """Get application settings."""
    return {"precision": 10}

@server.prompt()
def calculate(expression):
    # type: (str) -> list
    """Generate a calculation prompt."""
    return [
        {
            "role": "user",
            "content": {"type": "text", "text": "Please calculate: {}".format(expression)}
        }
    ]

if __name__ == "__main__":
    server.run(port=8000)

运行它：

python start_mcp.py

服务器启动并打印所有可用端点：

MCP Server 'my-calculator' v1.0.0 listening on 0.0.0.0:8000
  Tools: 2
  Resources: 1
  Prompts: 1
Endpoints:
  SSE transport:        http://0.0.0.0:8000/sse
  Streamable HTTP:      http://0.0.0.0:8000/mcp
  OpenAPI schema:       http://0.0.0.0:8000/openapi.json
  MCP discovery:        http://0.0.0.0:8000/.well-known/mcp.json
  Direct tool calls:    http://0.0.0.0:8000/tools/

______________________________________________________________________

服务器端点

端点方法描述
/GET服务器信息（名称、版本、工具/资源/提示计数）
/sseGETSSE传输--将响应作为服务器发送事件流式传输
/mcpGET流式HTTP——SSE流 endpoint 事件
/mcpPOST流式HTTP——接受JSON-RPC请求
/POSTJSON-RPC端点（与 /mcp 职位）
/tools/POST直接REST风格的工具调用（兼容OpenAPI）
/openapi.jsonGET自动生成的OpenAPI 3.1模式
/.well-known/mcp.jsonGETMCP发现文档

______________________________________________________________________

测试您的服务器

1.服务器信息

curl http://localhost:8000/
{
  "name": "my-calculator",
  "version": "1.0.0",
  "status": "running",
  "tools": 2,
  "resources": 1,
  "prompts": 1,
  "endpoints": {"sse": "/sse", "mcp": "/mcp", "openapi": "/openapi.json"}
}

2.MCP JSON-RPC（标准协议）

初始化：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{}}'
{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "protocolVersion": "2024-11-05",
    "serverInfo": {"name": "my-calculator", "version": "1.0.0"},
    "capabilities": {"tools": {}, "resources": {}, "prompts": {}, "logging": {}}
  }
}

列出工具：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/list","params":{}}'

调用工具：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"add","arguments":{"a":2,"b":3}}}'
{
  "jsonrpc": "2.0",
  "id": 3,
  "result": {
    "content": [{"type": "text", "text": "5.0"}],
    "isError": false
  }
}

列出资源：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":4,"method":"resources/list","params":{}}'

阅读资源：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":5,"method":"resources/read","params":{"uri":"config://settings"}}'

列表提示：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":6,"method":"prompts/list","params":{}}'

获取提示：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":7,"method":"prompts/get","params":{"name":"calculate","arguments":{"expression":"2+2"}}}'

发出砰的声响：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":8,"method":"ping","params":{}}'

3.直接工具调用（REST/OpenAPI风格）

无需JSON-RPC包装即可直接调用工具：

curl -X POST http://localhost:8000/tools/add \
  -H "Content-Type: application/json" \
  -d '{"a": 7, "b": 3}'
{"result": "10.0"}

4.苏格兰和南方能源公司运输

连接到SSE流以接收实时响应：

curl -N http://localhost:8000/sse

在单独的终端中，发送请求：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"ping","params":{}}'

SSE流将排放：

event: open
data: {}

event: message
data: {"jsonrpc":"2.0","id":1,"result":{}}

5.OpenAPI模式

curl http://localhost:8000/openapi.json

返回一个完整的OpenAPI 3.1模式，其中所有工具都作为POST端点公开。

6.MCP发现

curl http://localhost:8000/.well-known/mcp.json
{
  "mcpVersion": "2024-11-05",
  "serverInfo": {"name": "my-calculator", "version": "1.0.0"},
  "capabilities": {"tools": {}, "resources": {}, "prompts": {}, "logging": {}},
  "transports": [
    {"type": "sse", "endpoint": "/sse"},
    {"type": "streamable-http", "endpoint": "/mcp"}
  ]
}

7.Python测试客户端

import json
try:
    from http.client import HTTPConnection
except ImportError:
    from httplib import HTTPConnection

conn = HTTPConnection("localhost", 8000, timeout=5)

def call(method, params=None):
    body = json.dumps({
        "jsonrpc": "2.0",
        "id": 1,
        "method": method,
        "params": params or {}
    }).encode("utf-8")
    conn.request("POST", "/", body=body, headers={"Content-Type": "application/json"})
    resp = conn.getresponse()
    return json.loads(resp.read().decode("utf-8"))

# Initialize
print(call("initialize"))

# List tools
print(call("tools/list"))

# Call a tool
print(call("tools/call", {"name": "add", "arguments": {"a": 10, "b": 20}}))

# List resources
print(call("resources/list"))

# Read a resource
print(call("resources/read", {"uri": "config://settings"}))

# List prompts
print(call("prompts/list"))

# Get a prompt
print(call("prompts/get", {"name": "calculate", "arguments": {"expression": "2+2"}}))

conn.close()

______________________________________________________________________

api参考

MCP服务器

from nanohubmcp import MCPServer

server = MCPServer("my-server", version="1.0.0")
server.run(host="0.0.0.0", port=8000, path_prefix="")
参数类型默认值说明
namestr必填服务器名称
versionstr"1.0.0"服务器版本

server.run() 参数：

参数类型默认值说明
hoststr"0.0.0.0"要绑定的主机
portint8000要收听的端口
path_prefixstr""代理环境的URL前缀

@server.tool（）

将函数注册为MCP工具。函数名变为工具名，docstring变为描述。参数从函数签名中自动检测。

@server.tool()
def add(a, b):
    # type: (float, float) -> float
    """Add two numbers together."""
    return float(a) + float(b)

有选项：

@server.tool(name="custom_name", description="Custom description", tags={"math"})
def my_func(a, b):
    return a + b
参数类型默认值说明
namestr函数名工具名
descriptionstrdocstring工具说明
tagssetNone分类标签
meta 字典 None元数据字典
input_schemadict自动生成用于输入的JSON模式

@server.async_tool（）

将长时间运行的函数注册为异步MCP工具。服务器返回一个 job_id 立即而不是封锁。对于任何可能超过反向代理超时（通常为30-60秒）的工具，请使用此选项。

from nanohubmcp import MCPServer

server = MCPServer("my-server")

@server.async_tool()
def run_simulation(verilog_code: str, design_name: str) -> str:
    """Run a long RTL-to-GDSII flow. Can take up to 10 minutes."""
    # ... long-running work ...
    return result

召唤 run_simulation 立即返回作业ID：

{
  "status": "running",
  "job_id": "550e8400-e29b-41d4-a716-446655440000",
  "message": "Job started. Poll with get_job_result(job_id=\"...\")"
}

然后，客户端使用内置的 get_job_result 工具（在每台服务器上自动注册）：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"get_job_result","arguments":{"job_id":"550e8400-..."}}}'

退货 {"status": "running"} 在进行过程中，则：

{ "status": "done", "job_id": "...", "result": "..." }

或者在失败时： {"status": "error", "job_id": "...", "error": "..."}.

接受与相同的参数 @server.tool():

参数类型默认值说明
namestr函数名工具名
descriptionstrdocstring工具说明
tagssetNone分类标签
meta 字典 None元数据字典
input_schemadict自动生成用于输入的JSON模式

@server.resource（）

将函数注册为MCP资源。

@server.resource("config://calculator/settings")
def get_settings():
    """Get calculator settings."""
    return {"precision": 10}

MIME类型：

@server.resource("data://samples/temperatures", mime_type="application/json")
def temperature_data():
    """Monthly average temperatures."""
    return {"data": [2.1, 3.5, 7.2, 12.1]}
参数类型默认值说明
uristr必需资源URI
namestr函数名资源名
descriptionstrdocstring资源描述
mime_typestrNoneMIME内容类型
tagssetNone分类标签
meta 字典 None元数据字典

@server.promp（）

将函数注册为MCP提示模板。

@server.prompt()
def calculate(expression):
    # type: (str) -> list
    """Generate a calculation prompt."""
    return [
        {
            "role": "user",
            "content": {"type": "text", "text": "Please calculate: {}".format(expression)}
        }
    ]
参数类型默认值说明
namestr函数名提示名
descriptionstrdocstring提示描述
tagssetNone分类标签
meta 字典 None元数据字典

上下文

工具可以接收 Context 用于记录和进度报告的对象。添加a ctx （或 context)参数作为第一个参数：

from nanohubmcp import MCPServer, Context

server = MCPServer("my-server")

@server.tool()
def power(ctx, base, exponent):
    # type: (Context, float, float) -> float
    """Raise base to the power of exponent."""
    ctx.info("Computing {}^{}".format(base, exponent))
    ctx.report_progress(0.5, total=1.0, message="Computing...")
    return float(base) ** float(exponent)

上下文方法：

方法说明
ctx.debug(msg)日志调试消息
ctx.info(msg)日志信息消息
ctx.warning(msg)日志警告消息
ctx.error(msg)日志错误消息
ctx.report_progress(progress, total, message)向客户报告进度

______________________________________________________________________

示例

学生用函数计算器

一个具有算术运算、设置资源和计算提示的基本计算器。

from nanohubmcp import MCPServer, Context

server = MCPServer("simple-calculator", version="1.0.0")

@server.tool()
def add(a, b):
    # type: (float, float) -> float
    """Add two numbers together."""
    return float(a) + float(b)

@server.tool(tags={"math", "advanced"})
def power(ctx, base, exponent):
    # type: (Context, float, float) -> float
    """Raise base to the power of exponent. Demonstrates Context usage."""
    ctx.info("Computing {}^{}".format(base, exponent))
    return float(base) ** float(exponent)

@server.tool()
def subtract(a, b):
    # type: (float, float) -> float
    """Subtract b from a."""
    return float(a) - float(b)

@server.tool()
def multiply(a, b):
    # type: (float, float) -> float
    """Multiply two numbers."""
    return float(a) * float(b)

@server.tool()
def divide(a, b):
    # type: (float, float) -> float
    """Divide a by b."""
    if float(b) == 0:
        raise ValueError("Cannot divide by zero")
    return float(a) / float(b)

@server.resource("config://calculator/settings")
def get_settings():
    """Get calculator settings."""
    return {
        "precision": 10,
        "max_value": 1e308,
        "supported_operations": ["add", "subtract", "multiply", "divide", "power"]
    }

@server.prompt()
def calculate(expression):
    # type: (str) -> list
    """Generate a calculation prompt."""
    return [
        {
            "role": "user",
            "content": {"type": "text", "text": "Please calculate: {}".format(expression)}
        }
    ]

if __name__ == "__main__":
    server.run(port=8000)

测试一下：

# Call add
curl -X POST http://localhost:8000/tools/add \
  -H "Content-Type: application/json" -d '{"a": 2, "b": 3}'

# Call power
curl -X POST http://localhost:8000/tools/power \
  -H "Content-Type: application/json" -d '{"base": 2, "exponent": 10}'

# Call divide (error case)
curl -X POST http://localhost:8000/tools/divide \
  -H "Content-Type: application/json" -d '{"a": 1, "b": 0}'

查看完整来源： examples/simple/start_mcp.py

______________________________________________________________________

数据分析

用于数据探索的带有样本数据集的统计分析工具。

import math
from nanohubmcp import MCPServer

server = MCPServer("data-analysis", version="1.0.0")

@server.tool()
def descriptive_stats(data):
    # type: (str) -> dict
    """
    Calculate descriptive statistics for a dataset.

    Args:
        data: Comma-separated list of numeric values (e.g., "1,2,3,4,5")
    """
    values = [float(x.strip()) for x in data.split(",")]
    n = len(values)
    sorted_data = sorted(values)
    mean = sum(values) / n

    if n % 2 == 0:
        median = (sorted_data[n//2 - 1] + sorted_data[n//2]) / 2
    else:
        median = sorted_data[n//2]

    variance = sum((x - mean) ** 2 for x in values) / n
    std = math.sqrt(variance)

    return {
        "count": n, "mean": round(mean, 6), "median": round(median, 6),
        "min": min(values), "max": max(values), "std": round(std, 6)
    }

@server.tool()
def correlation(x_data, y_data):
    # type: (str, str) -> dict
    """Calculate Pearson correlation coefficient between two datasets."""
    x = [float(v.strip()) for v in x_data.split(",")]
    y = [float(v.strip()) for v in y_data.split(",")]
    # ... (see full source for implementation)

@server.tool()
def linear_regression(x_data, y_data):
    # type: (str, str) -> dict
    """Perform simple linear regression (y = mx + b)."""
    # ...

@server.tool()
def normalize(data, method="minmax"):
    # type: (str, str) -> dict
    """Normalize a dataset using 'minmax' or 'zscore' method."""
    # ...

@server.resource("data://samples/temperatures", mime_type="application/json")
def temperature_data():
    """Monthly average temperatures (Celsius) for a year."""
    return {
        "data": [2.1, 3.5, 7.2, 12.1, 17.3, 21.5, 24.2, 23.8, 19.4, 13.2, 7.1, 3.2],
        "labels": ["Jan", "Feb", "Mar", "Apr", "May", "Jun",
                   "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"]
    }

@server.resource("data://samples/scatter", mime_type="application/json")
def scatter_data():
    """Sample data for scatter plot / correlation analysis."""
    return {
        "x": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
        "y": [52, 58, 65, 68, 72, 78, 82, 85, 90, 95]
    }

@server.prompt()
def analyze_data(data):
    # type: (str) -> list
    """Generate a prompt to analyze a dataset."""
    return [{"role": "user", "content": {"type": "text", "text": "Please analyze: {}".format(data)}}]

if __name__ == "__main__":
    server.run(port=8000)

测试一下：

# Descriptive statistics
curl -X POST http://localhost:8000/tools/descriptive_stats \
  -H "Content-Type: application/json" \
  -d '{"data": "10,20,30,40,50"}'

# Correlation
curl -X POST http://localhost:8000/tools/correlation \
  -H "Content-Type: application/json" \
  -d '{"x_data": "1,2,3,4,5", "y_data": "2,4,6,8,10"}'

# Linear regression
curl -X POST http://localhost:8000/tools/linear_regression \
  -H "Content-Type: application/json" \
  -d '{"x_data": "1,2,3,4,5", "y_data": "2.1,3.9,6.2,7.8,10.1"}'

# Normalize
curl -X POST http://localhost:8000/tools/normalize \
  -H "Content-Type: application/json" \
  -d '{"data": "10,20,30,40,50", "method": "zscore"}'

# Read temperature dataset
curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"resources/read","params":{"uri":"data://samples/temperatures"}}'

查看完整来源： 示例/数据分析/start_mcp.py

______________________________________________________________________

物理模拟器

以物理常数为资源的物理模拟工具。

import math
from nanohubmcp import MCPServer, Context

server = MCPServer("physics-simulator", version="1.0.0")

GRAVITY = 9.81
SPEED_OF_LIGHT = 299792458

@server.tool()
def projectile_motion(v0, angle, h0=0):
    # type: (float, float, float) -> dict
    """
    Calculate projectile motion parameters.

    Args:
        v0: Initial velocity (m/s)
        angle: Launch angle (degrees)
        h0: Initial height (m), default 0
    """
    v0 = float(v0)
    angle_rad = math.radians(float(angle))
    vx = v0 * math.cos(angle_rad)
    vy = v0 * math.sin(angle_rad)
    t_flight = (vy + math.sqrt(vy**2 + 2 * GRAVITY * float(h0))) / GRAVITY
    return {
        "range": round(vx * t_flight, 3),
        "max_height": round(float(h0) + vy**2 / (2 * GRAVITY), 3),
        "time_of_flight": round(t_flight, 3)
    }

@server.tool()
def harmonic_oscillator(mass, spring_constant, amplitude, time):
    # type: (float, float, float, float) -> dict
    """Calculate simple harmonic motion parameters."""
    # ...

@server.tool()
def wave_properties(frequency, wavelength=None, medium_speed=None):
    # type: (float, float, float) -> dict
    """Calculate wave properties (period, speed, wavelength, photon energy)."""
    # ...

@server.tool()
def ideal_gas(pressure=None, volume=None, n_moles=None, temperature=None):
    # type: (float, float, float, float) -> dict
    """Ideal gas law calculator (PV = nRT). Provide 3 of 4 variables."""
    # ...

@server.tool(tags={"advanced"})
def relativistic_energy(ctx, rest_mass, velocity):
    # type: (Context, float, float) -> dict
    """Calculate relativistic energy and momentum."""
    ctx.info("Calculating relativistic properties for v = {} m/s".format(velocity))
    # ...

@server.resource("constants://physics", mime_type="application/json")
def physical_constants():
    """Fundamental physical constants."""
    return {
        "speed_of_light": {"value": 299792458, "unit": "m/s"},
        "gravitational_acceleration": {"value": 9.81, "unit": "m/s^2"},
        "planck_constant": {"value": 6.62607015e-34, "unit": "J*s"},
        "boltzmann_constant": {"value": 1.380649e-23, "unit": "J/K"}
    }

@server.prompt()
def physics_problem(problem_description):
    # type: (str) -> list
    """Generate a prompt to solve a physics problem."""
    return [{"role": "user", "content": {"type": "text", "text": "Solve: {}".format(problem_description)}}]

if __name__ == "__main__":
    server.run(port=8000)

测试一下：

# Projectile motion (50 m/s at 45 degrees)
curl -X POST http://localhost:8000/tools/projectile_motion \
  -H "Content-Type: application/json" \
  -d '{"v0": 50, "angle": 45}'

# Harmonic oscillator
curl -X POST http://localhost:8000/tools/harmonic_oscillator \
  -H "Content-Type: application/json" \
  -d '{"mass": 0.5, "spring_constant": 20, "amplitude": 0.1, "time": 1.0}'

# Ideal gas law (find pressure given V, n, T)
curl -X POST http://localhost:8000/tools/ideal_gas \
  -H "Content-Type: application/json" \
  -d '{"volume": 0.0224, "n_moles": 1, "temperature": 273.15}'

# Read physical constants
curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"resources/read","params":{"uri":"constants://physics"}}'

查看完整来源： 示例/模拟器/start_mcp.py

______________________________________________________________________

在nanoHUB上运行

在nanoHUB上，使用 start_mcp CLI命令：

start_mcp --app start_mcp.py

在conda环境中：

start_mcp --app start_mcp.py --python-env AIIDA

它在nanoHUB上的工作原理：

  1. CLI检测nanoHUB环境(SESSION, SESSIONDIR 变量）
  2. 它寻找一个可用的 wrwroxy 使用反向代理 use 命令
  3. 如果发现wrwrwroxy：MCP在端口8001上运行，wrwrwroxy代理端口8000
  4. 如果未找到wrwrwroxy：MCP直接在带有weber路径前缀的端口8000上运行

启动时打印代理URL：

Proxy URL : https://proxy.nanohub.org/weber/{session}/{cookie}/{port}/
MCP Server ready! Access it at: https://proxy.nanohub.org/weber/...
SSE endpoint: https://proxy.nanohub.org/weber/.../sse

通过nanoHUB代理进行测试：

PROXY_URL="https://proxy.nanohub.org/weber/{session}/{cookie}/{port}"
COOKIE="weber-auth-nanohub-org={session}%3A{auth_token}"

# Server info
curl -b "$COOKIE" "$PROXY_URL/"

# Initialize
curl -b "$COOKIE" -X POST "$PROXY_URL/" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{}}'

# Call a tool
curl -b "$COOKIE" -X POST "$PROXY_URL/" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"add","arguments":{"a":2,"b":3}}}'

______________________________________________________________________

运行测试

测试套件在后台线程中启动服务器，并执行所有端点：

pip install pytest
pytest tests/
tests/test_mcp_server.py ...................... [100%]
22 passed

______________________________________________________________________

MCP协议参考

支持的JSON-RPC方法

方法说明
initialize初始化连接，返回协议版本和功能
ping健康检查，退货 {}
tools/list列出所有已注册的带有模式的工具
tools/call按名称和参数调用工具
resources/list列出所有已注册的资源
resources/read按URI读取资源
prompts/list列出所有已注册的提示
prompts/get按名称和参数获取提示

通知

JSON-RPC请求没有 id 字段被视为通知并接收 202 Accepted 响应：

curl -X POST http://localhost:8000/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"initialized","params":{}}'
{"status": "accepted"}

错误处理

刀具错误返回 isError: true 在MCP响应中：

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "content": [{"type": "text", "text": "Cannot divide by zero"}],
    "isError": true
  }
}

未知方法返回JSON-RPC错误：

{
  "jsonrpc": "2.0",
  "id": 1,
  "error": {"code": -32601, "message": "Method not found: unknown/method"}
}

直接工具调用错误(/tools/)返回HTTP 500：

{"error": "Cannot divide by zero"}

______________________________________________________________________

许可证

MIT许可证-有关详细信息，请参阅许可证文件。

目录标签

目录标签

服务器开发Python工具集成Python库本地部署MCP协议科学计算

接入字段

传输方式（transport，传输协议）

stdio

鉴权方式（authType，认证方式）

session

工具数量（toolCount，工具数）

2

资源数量（resourceCount，资源数）

0

提示词数量（promptCount，提示词数）

0

权限和风险

stdiosession部署方式未说明

接入前请确认传输方式、认证方式和部署位置，并根据实际工具能力限制访问范围。

安装前确认

不要直接授予不必要的文件、网络或账号权限；先核对安装命令和配置内容。

来源信息

nanohub-mcp logo

MCP World

mcpworld.com

/detail/3699b040e59f1c0695bdefa3bab87609

一个零依赖的Python库，用于创建与nanoHUB/HubZero工具基础设施集成的Model Context Protocol (MCP)服务器。

继续浏览同类 MCP

更多AI代理 MCP（18,706）返回 MCP 目录

上一个 MCP

Nano Currency MCP Server

Nano货币MCP服务器为MCP兼容客户端（包括AI代理）提供通过Nano节点RPC发送Nano货币和检索账户及区块信息的能力，并支持创建x402兼容头以实现API/资源的Nano支付。

下一个 MCP

Napkin 74fe7a8f

A shared whiteboard for you and your AI agent I wanted my AI agent and me to be able to point at the same thing. Any MCP-capable agent can read the canvas, draw on it, drop thought bubbles, animate elements, and react when you sketch something.