My Prompts MCP
Lightweight MCP server for providing user prompts defined in Markdown with argument support.
README
My Prompts MCP
Lightweight MCP server that serves user prompts defined in Markdown files, with support for arguments.
Keep your prompt .md files in a repository — a dedicated git repo or a folder in an existing one —
and connect that directory to your AI client via this MCP server. This way your prompts are versioned,
shareable, and easy to update across machines and team members.
Example of My Prompts MCP usage with Cherry Studio:
https://github.com/user-attachments/assets/32dfa36d-657f-4b9b-8086-c7a00ffe3185
[!IMPORTANT] This project is developed and maintained by Sergei Predvoditelev. Community support helps keep the project actively developed and well maintained. You can support the project using the following services:
Thank you for your support ❤️
General Usage
Installation
Download the binary for your platform from GitHub Releases.
Create a prompt
Create a .md file in your prompts directory, for example greet.md:
---
description: Greet a person by name
arguments:
- name: name
description: The person's name
required: true
---
Please greet {{name}} in a friendly and professional way.
Connecting to an MCP client
Add the following to your MCP client configuration (stdio transport):
{
"mcpServers": {
"my-prompts-mcp": {
"command": "/path/to/my-prompts-mcp",
"args": ["--path=/path/to/prompts"]
}
}
}
<details> <summary>Claude Desktop</summary>
Edit the claude_desktop_config.json file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
Add the snippet above to the file, then restart Claude Desktop.
</details>
<details> <summary>Cherry Studio</summary>
Open Settings → MCP Servers, click Add, and fill in the server command and arguments from the snippet above.
</details>
Documentation
Prompt file format
Each .md file in the configured directory is exposed as a prompt. The file name without the .md extension is used as
the prompt name by default.
Front matter fields
Optional YAML front matter at the top of the file controls how the prompt is presented to the client.
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | No | Prompt name. Defaults to the filename without extension. |
title |
string | No | Human-readable title. |
description |
string | No | Short description shown to the AI client. |
arguments |
list | No | List of arguments (see below). |
Arguments
Arguments can be defined in two forms:
Simple (name only):
arguments:
- arg_name
Full object:
arguments:
- name: arg_name
description: What this argument means
required: true
Each argument field:
| Field | Type | Required | Description |
|---|---|---|---|
name |
string | Yes | Argument name. |
description |
string | No | Description of the argument. |
required |
boolean | No | Whether the argument is required. Defaults to false. |
[!IMPORTANT] Prompts without arguments or with one argument are supported by almost all MCP clients. Prompts with two or more arguments are not universally supported. Known clients that do support multiple arguments:
Placeholders
Use {{arg_name}} in the prompt body. Placeholders are substituted with the argument values provided at request time.
Example prompt file
---
name: name-generator
title: Name Generator
description: Generate a name for a product, project, or company
arguments:
- name: description
description: What needs to be named (product, project, company, etc.)
required: true
- name: style
description: Naming style (e.g. minimalist, creative, technical, playful)
required: false
---
Generate 10 name ideas for: {{description}}.
{{style}}
Requirements for the names:
- Easy to remember and pronounce
- Suitable for use as a domain name
- Unique and distinctive
For each name provide a one-line explanation of why it works.
License
The "My Prompts MCP" is free software. It is released under the terms of the BSD License.
Please see LICENSE for more information.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。