Model Context Protocol (MCP) Server
使用 LangChain / Python 的 MCP 客户端
这个简单的模型上下文协议 (MCP)客户端演示了 LangChain ReAct Agent 对 MCP 服务器工具的使用。
它利用了langchain_mcp_tools中的实用函数convert_mcp_to_langchain_tools() 。
此函数处理指定的多个 MCP 服务器的并行初始化,并将它们可用的工具转换为与 LangChain 兼容的工具列表 ( List[BaseTool] )。
目前支持 Anthropic、OpenAI 和 Groq 的 LLM。
此 MCP 客户端的 TypeScript 版本可在此处获取
先决条件
Python 3.11+
Related MCP server: Just Prompt
设置
安装依赖项:
make install设置 API 密钥:
cp .env.template .env根据需要更新
.env。.gitignore配置为忽略.env以防止意外提交凭据。
根据需要配置 LLM 和 MCP 服务器设置
llm_mcp_config.json5。MCP 服务器的配置文件格式遵循与Claude for Desktop相同的结构,但有一点不同:键名
mcpServers已更改为mcp_servers,以遵循 JSON 配置文件中常用的 snake_case 约定。文件格式为JSON5 ,其中允许使用注释和尾随逗号。
该格式进一步扩展,用相应环境变量的值替换
${...}符号。将所有凭证和私人信息保存在
.env文件中,并根据需要使用${...}符号引用它们。
用法
运行应用程序:
make start第一次运行需要一段时间。
以详细模式运行:
make start-v查看命令行选项:
make start-h在提示符下,您只需按 Enter 即可使用执行 MCP 服务器工具调用的示例查询。
可以在llm_mcp_config.json5中配置示例查询
Available Tools
2 toolsget-alertsC
Get weather alerts for a US state
| Name | Required | Description | Default |
|---|---|---|---|
| state | Yes | Two-letter US state code (e.g. CA, NY) |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description must carry full behavioral transparency. It only states the basic action without disclosing traits like read-only nature, authentication requirements, rate limits, or what type of alerts are returned. This is insufficient for an agent to understand the tool's behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single sentence, achieving conciseness. It is appropriately front-loaded with the verb and resource. However, it lacks structure such as prerequisites or return format, but it remains efficient for its length.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simplicity (one required parameter, no output schema, no annotations), the description should provide more context, such as what the alerts contain or that it is a read operation. The minimal description leaves gaps for an agent to understand the tool's full context.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema has 100% description coverage for the 'state' parameter, so the baseline is 3. The description does not add any additional meaning beyond the schema; it simply restates the parameter's role without further context.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Get') and resource ('weather alerts') with a specific scope ('for a US state'). It distinguishes itself from the sibling 'get-forecast' by focusing on alerts, not forecasts, though it does not explicitly mention the sibling.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives like 'get-forecast'. The description does not include any prerequisites, context, or exclusions, leaving the agent to infer usage from tool names alone.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
get-forecastB
Get weather forecast for a location in the US
| Name | Required | Description | Default |
|---|---|---|---|
| latitude | Yes | Latitude of the location | |
| longitude | Yes | Longitude of the location |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, and the description does not disclose any behavioral traits such as data source, update frequency, or limitations of the forecast. The description only states the basic purpose.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single concise sentence that immediately communicates the tool's purpose. No unnecessary words.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple tool with two parameters and no output schema, the description is minimal but covers the core purpose. However, it lacks usage guidance and behavioral details that would help an AI agent use it correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% with both parameters described. The description adds no additional meaning beyond the schema; the mention of 'in the US' is a location constraint but not parameter-specific. Baseline of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action 'Get', the resource 'weather forecast', and the scope 'for a location in the US'. It distinguishes from the sibling tool 'get-alerts' which likely deals with alerts.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance on when to use this tool versus alternatives. The description does not indicate when to prefer this over 'get-alerts' or any other context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
2 tool updates
v0.3.2- First observed
get-alerts - First observed
get-forecast
TDQS
Scored across 2 tools
The tools get-alerts and get-forecast have clearly distinct purposes: one provides weather alerts, the other gives forecasts. No overlap.
Both tool names follow the consistent pattern 'get-<resource>', using lowercase and hyphens, which is predictable.
With only 2 tools, the server feels too sparse for a weather domain. Typically, more operations like current conditions or radar would be expected.
The tool set only covers alerts and forecasts, missing common weather operations like current conditions, location search, or severe weather warnings, resulting in significant gaps.
Maintenance
Related MCP Connectors
AI routing, memory, guardrails, and governance. Routes across Claude, GPT, Gemini.
Hosted MCP server for LLM cost estimation, model comparison, and budget-aware routing.
MCP server unifying ERPs, CRMs, APIs and knowledge base for Claude, ChatGPT and Gemini.
Furgonetka MCP Server is an extension for LLMs (such as Claude) that integrates AI assistants with Poland's most popular courier brokerage platform. The server enables models to interact directly with services from various couriers (including InPost, DPD, DHL, UPS, and Poczta Polska) through a single, unified interface. With this integration, your AI stops just "writing about logistics" and starts actually managing it.
Related MCP Servers
- FlicenseNot gradedqualityFmaintenanceThis server provides an API to query Large Language Models using context from local files, supporting various models and file types for context-aware responses.1-
- FlicenseBqualityDmaintenanceA lightweight MCP server that provides a unified interface to various LLM providers including OpenAI, Anthropic, Google Gemini, Groq, DeepSeek, and Ollama.6738-
- AlicenseNot gradedqualityDmaintenanceA simple server that acts as a Master Control Program (MCP) for unified interaction with OpenAI and Anthropic (Claude) AI models through a single API endpoint.23 npmMIT
- FlicenseNot gradedqualityDmaintenanceA unified API server that enables interaction with multiple AI model providers like Anthropic and OpenAI through a consistent interface, supporting chat completions, tool calling, and context handling.-