tae4an/mcp-prompt-manager

Developer tools

Эта страница переведена автоматически. Технические детали смотрите в оригинале. Открыть оригинал

5 stars 3 forks Качество 69 Тренд 13MIT

Repository Install

Local Prompt Management System(Model Context Protocol Server)

Обзор

This project is officially certified by MCP Review. MCP (Model Context Protocol) Prompt Manager is a server that enables AI models like Claude to access local prompt files. It provides functionality for creating, retrieving, updating, and deleting prompts, allowing efficient management of frequently used prompts. - List all prompts - Retrieve specific prompt content - Create new prompts - Update prompt content - Delete prompts - : Advanced fuzzy search with multiple algorithms (Levenshtein, Jaro-Winkler, n-gram similarity) - : Organize prompts with categories and tags - : Use variable substitution with {{variable}} syntax and advanced conditionals - : Built-in template library with 12 professional templates across 5 categories - : Mark frequently used prompts as favorites - : Automatic metadata tracking for enhanced organization - : Complete version control with history tracking, diff comparison, and rollback capabilities - : Backup and restore prompts with JSON format, including...

README

Показан фрагмент оригинального README. Открыть оригинал.

MCP Prompt Manager

Certified by MCP Review

This project is officially certified by MCP Review.

MCP (Model Context Protocol) Prompt Manager is a server that enables AI models like Claude to access local prompt files. It provides functionality for creating, retrieving, updating, and deleting prompts, allowing efficient management of frequently used prompts.

Key Features

Core Features

List all prompts
Retrieve specific prompt content
Create new prompts
Update prompt content
Delete prompts

Advanced Features

Intelligent Search: Advanced fuzzy search with multiple algorithms (Levenshtein, Jaro-Winkler, n-gram similarity)
Category & Tag System: Organize prompts with categories and tags
Template Processing: Use variable substitution with {{variable}} syntax and advanced conditionals
Template Library: Built-in template library with 12 professional templates across 5 categories
Favorites Management: Mark frequently used prompts as favorites
Metadata Management: Automatic metadata tracking for enhanced organization
Version Management: Complete version control with history tracking, diff comparison, and rollback capabilities
Import/Export System: Backup and restore prompts with JSON format, including metadata and version history
Security & Validation: Comprehensive input sanitization, rate limiting, and error handling
Caching System: Intelligent caching for improved performance
Structured Logging: Advanced logging with multiple levels and file output

Installation

Prerequisites

Node.js v18 or higher
npm

Installation Steps

Clone the repository

git clone https://github.com/Tae4an/mcp-prompt-manager.git
cd mcp-prompt-manager

Install dependencies
```
npm install
```
Grant execution permissions
```
chmod +x server.js
```

Containerized Run

Run with Docker

docker build -t mcp-prompt-manager:local .
docker run --rm \
  -e NODE_ENV=production \
  -e LOG_DIR=/var/log/mcp \
  -e PROMPTS_DIR=/data/prompts \
  -v $(pwd)/prompts:/data/prompts \
  -v $(pwd)/logs:/var/log/mcp \
  mcp-prompt-manager:local

Run with Docker Compose

docker compose up -d --build

Environment variables

PROMPTS_DIR: Directory to store prompts (default: project prompts)
LOG_DIR: Directory for file logging (default: ./logs)
Cache TTL/size:
- FILE_CACHE_TTL, FILE_CACHE_MAX_SIZE
- SEARCH_CACHE_TTL, SEARCH_CACHE_MAX_SIZE
- METADATA_CACHE_TTL, METADATA_CACHE_MAX_SIZE
- TEMPLATE_CACHE_TTL, TEMPLATE_CACHE_MAX_SIZE
Rate limit presets:
- RATE_LIMIT_STANDARD_WINDOW_MS, RATE_LIMIT_STANDARD_MAX
- RATE_LIMIT_STRICT_WINDOW_MS, RATE_LIMIT_STRICT_MAX
- RATE_LIMIT_LENIENT_WINDOW_MS, RATE_LIMIT_LENIENT_MAX

Открыть полный README на GitHub

Установка

docker run --rm \

Новичок в MCP? Начните с руководства на mcpskill.org →

Конфигурация

{
     "mcpServers": {
       "promptManager": {
         "command": "node",
         "args": ["absolute_path_to_cloned_repository/server.js"]
       }
     }
   }