Specif-ai MCP Server
A CLI tool that runs an MCP server over stdio for Specif-ai, enabling interaction with specification documents like BRDs, PRDs, and user stories through a set of structured tools.
Tools
set-project-path
Set the project path and reload the solution, use this tool only when we not automatically able to infer the project path or asked by the user or us., we will try to auto infer it from the environment first.
get-brds
Get Business Requirement Documents for this project
get-prds
Get Product Requirement Documents for this project
get-nfrs
Get Non-Functional Requirement Documents for this project
get-uirs
Get User Interface Requirement Documents for this project
get-bps
Get Business Process Documents for this project
get-user-stories
Get User Stories for a particular PRD
get-tasks
Get Tasks for a particular User Story
get-task
Get a Task for a particular User Story in a particular PRD
README
@vj-presidio/specif-ai-mcp-server
A CLI tool that runs an MCP (Model Context Protocol) server over stdio for Specif-ai.
Installation & Usage Overview
graph TD
A[Installation Options]
A --> B[Direct Binary Installation]
A --> C[Package Manager Installation]
B --> D[Unix Script]
B --> E[Windows PowerShell]
C --> F[NPM Global]
C --> G[Bun Global]
B --> H1[When to Use Binary Installation]
H1 --> I1[System-wide installation]
H1 --> I2[No Node.js required]
H1 --> I3[Minimal dependencies]
C --> H2[When to Use Package Manager]
H2 --> J1[Already using Node.js/Bun]
H2 --> J2[Need automatic updates]
H2 --> J3[Project-specific versions]
K[Usage Methods]
K --> L[Direct CLI]
K --> M[NPX]
K --> N[BunX]
L --> O1[Installed globally]
L --> O2[Fastest startup]
M --> P1[No installation needed]
M --> P2[Version flexibility]
N --> Q1[Bun runtime users]
N --> Q2[Better performance]
Installation
Direct Binary Installation (Recommended)
You can install the binary directly using our installation script:
# Unix (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/install.sh | sh
# Install specific version
curl -fsSL https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/install.sh | sh -s -- -v 1.2.3
# Windows (PowerShell)
iwr -useb https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/install.ps1 | iex
# Install specific version
iwr -useb https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/install.ps1 | iex -v 1.2.3
or manually download the binary for your platform from the Releases page.
Package Manager Installation
You can install globally using npm:
# Latest version
npm install -g @vj-presidio/specif-ai-mcp-server@latest
# Specific version
npm install -g @vj-presidio/specif-ai-mcp-server@1.2.3
Or using bun:
# Latest version
bun install -g @vj-presidio/specif-ai-mcp-server@latest
# Specific version
bun install -g @vj-presidio/specif-ai-mcp-server@1.2.3
Updates
To check for updates:
# Unix (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.sh | sh -s -- -c
# Windows (PowerShell)
iwr -useb https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.ps1 | iex -c
To update to the latest version:
# Unix (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.sh | sh
# Windows (PowerShell)
iwr -useb https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.ps1 | iex
To update to a specific version:
# Unix (macOS/Linux)
curl -fsSL https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.sh | sh -s -- -v 1.2.3
# Windows (PowerShell)
iwr -useb https://raw.githubusercontent.com/vj-presidio/specif-ai-mcp-server/main/update.ps1 | iex -v 1.2.3
Example MCP Client Configuration
with npx with latest version:
{
"specif-ai": {
"command": "npx",
"args": ["--yes", "@vj-presidio/specif-ai-mcp-server@latest"],
"disabled": false,
"autoApprove": []
}
}
with npx with specific version:
{
"specif-ai": {
"command": "npx",
"args": ["--yes", "@vj-presidio/specif-ai-mcp-server@1.2.3"],
"disabled": false,
"autoApprove": []
}
}
with bunx with latest version:
{
"specif-ai": {
"command": "bunx",
"args": ["@vj-presidio/specif-ai-mcp-server@latest"],
"disabled": false,
"autoApprove": []
}
}
with bunx with specific version:
{
"specif-ai": {
"command": "bunx",
"args": ["@vj-presidio/specif-ai-mcp-server@1.2.3"],
"disabled": false,
"autoApprove": []
}
}
with direct binary or package manager global installation:
{
"specif-ai": {
"command": "specif-ai-mcp-server",
"args": [],
"disabled": false,
"autoApprove": []
}
}
Options
--help,-h: Display help information--version: Display version information
Specifai MCP integration with popular IDE and extension
See the setup instructions for each
Setting Project Path
Once the server is running, you can set the project path using the set-project-path tool. This tool accepts a path to the directory containing your specification files. After setting the path, the server will load all documents from that directory.
Example tool usage:
{
"name": "set-project-path",
"arguments": {
"path": "./path/to/project"
}
}
Available Tools
The server provides several tools for interacting with your specification documents:
| Tool Name | Description |
|---|---|
set-project-path |
Set or change the project directory path |
get-brds |
Get Business Requirement Documents |
get-prds |
Get Product Requirement Documents |
get-nfrs |
Get Non-Functional Requirements |
get-uirs |
Get User Interface Requirements |
get-bps |
Get Business Process Documents |
get-user-stories |
Get User Stories for a specific PRD |
get-tasks |
Get Tasks for a specific User Story |
get-task |
Get details of a specific Task |
Requirements
For binary installation:
- curl (Unix) or PowerShell (Windows)
- sudo access (Unix, for system-wide installation)
For package manager installation:
- Node.js >= 16.0.0
- Bun >= 1.0.0 (if using Bun runtime)
License
MIT
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。