Content
# EASP - Enterprise AI Service Platform (Open-Source Core)
[](https://go.dev/)
[](https://www.python.org/)
[](https://www.typescriptlang.org/)
[](LICENSE)
[](https://docs.docker.com/)
**EASP (Enterprise AI Service Platform)** open-source core - includes MCP protocol implementation, MCP Server framework, and embedded AI assistant SDK.
> 🌐 Official Website: [jindiyun.com](https://www.jindiyun.com) | Commercial Version: [easp.jindiyun.com](https://easp.jindiyun.com)
---
## 📦 Repository Structure
```
easp/
├── internal/
│ ├── mcp/ # Go MCP protocol implementation
│ │ ├── protocol.go # MCP protocol core (JSON-RPC 2.0)
│ │ ├── client.go # MCP Client (SSE + Streamable HTTP)
│ │ ├── server.go # MCP Server
│ │ ├── proxy.go # MCP proxy (transparent/ governance)
│ │ ├── curl_import.go # cURL → MCP tool import
│ │ └── builtin_governance.go # Built-in governance tool
│ └── openapi/
│ └── parser.go # OpenAPI → MCP conversion
├── cmd/
│ ├── mcp-test/ # MCP test tool (CLI)
│ └── mcp-e2e/ # MCP end-to-end test
├── mcp-server/ # Python MCP Server framework
│ ├── app/
│ │ ├── mcp_server.py # MCP Server implementation
│ │ ├── main.py # Entry point
│ │ └── config.py # Configuration
│ ├── requirements.txt
│ ├── mcp_config.json # Example configuration
│ └── start.sh / restart.sh / stop.sh
├── frontend/ # Embedded AI assistant SDK
│ ├── public/embed/
│ │ ├── assistant.js # Compiled output (5-line code integration)
│ │ └── assistant-frame.html # iframe mode
│ └── src/embed/
│ └── assistant-sdk.ts # TypeScript source code
├── docs/ # Documentation
├── migrations/ # Database migration (SSO example)
├── Dockerfile # Go MCP Tools container image
├── docker-compose.yml # One-click deployment of MCP Server
├── LICENSE # AGPL v3
└── README.md
```
---
## 🚀 Quick Start
### Go MCP Protocol Library
```bash
go get github.com/qinzhongyi249-rgb/easp
```
```go
import "github.com/qinzhongyi249-rgb/easp/internal/mcp"
// Create MCP Client
client := mcp.NewClient("http://localhost:9000/sse")
tools, _ := client.ListTools()
```
### Python MCP Server
```bash
cd mcp-server
pip install -r requirements.txt
bash start.sh
# MCP SSE endpoint: http://localhost:9000/sse
```
### Embedded AI Assistant SDK
```html
<script src="https://easp.jindiyun.com/embed/assistant.js"></script>
<script>
EASPAssistant.init({
appId: "your-app-id",
baseUrl: "https://easp.jindiyun.com"
});
</script>
```
---
## 🐳 Docker Deployment
```bash
# Start MCP Server + test tool
docker compose up -d mcp-server
# Run MCP test
docker compose --profile test run --rm mcp-test
# Verification
curl http://localhost:9000/sse
```
| Service | Port | Description |
|------|------|------|
| `mcp-server` | 9000 | Python MCP Server (SSE) |
| `mcp-test` | - | Go MCP test (one-time) |
---
## 📄 License
Copyright © 2026 Beijing Jindiyun Technology Co., Ltd.
This project uses [GNU Affero General Public License v3.0](LICENSE).
For commercial licensing, please contact [jindiyun.com](https://www.jindiyun.com).
## Community
Scan the QR code to join the Feishu group:
<img src="assets/feishu-qr.jpg" width="200" alt="Feishu discussion group">
---
**Beijing Jindiyun Technology Co., Ltd.** | Jing ICP No. 2026038568-1
