mcpField
T
Tiangong AI MCP

Tiangong AI MCP

Local
github.com/linancnOther4Updated 2w ago
Repository

About Tiangong AI MCP

TianGong AI MCP is an MCP server for education-resource search, with a Streamable HTTP option. Its current source-built version no longer provides the retired education knowledge graph tool.

Key features

  1. 01Education-resource search
  2. 02Streamable HTTP server

Use cases

  1. 01Search for education resources
  2. 02Connect through Streamable HTTP

MCP support

Provides an MCP server with Streamable HTTP support. The current server includes `Search_Edu_Tool`; the retired `Search_Edu_Graph_Tool` is no longer advertised.

#tiangong-ai#mcp#ai-protocol

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "tiangong-ai": {
      "command": "npx",
      "args": [
        "-y",
        "dotenv"
      ]
    }
  }
}
Run commandnpx -y dotenv

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmdotenv

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

docType: guide scope: repo status: current authoritative: true owner: mcp language: en whenToUse: 'When installing, configuring, running, or integrating the TianGong AI MCP server.' whenToUpdate: 'When tool behavior, auth, environment variables, package metadata, build commands, or server startup changes.' checkPaths:

  • AGENTS.md
  • .docpact/config.yaml
  • package.json
  • Dockerfile
  • .env.example
  • mcp_config.json
  • test.http
  • .github/prompts/**
  • README_CN.md
  • DEV_EN.md
  • DEV_CN.md
  • src/** lastReviewedAt: 2026-10-12 lastReviewedCommit: aabaceef96a38b0d23394e2da771b4fc2a213be0

TianGong-AI-MCP

GitHub organization migration / GitHub 组织迁移

This original repository now belongs to the tiangong-ai organization, with its repository identity and history retained. The CLI package @tiangong-ai/cli and command tiangong-ai are unchanged. Wiki is now published as @tiangong-ai/wiki; its commands remain unchanged. See the migration and upgrade notes.

该仓库已迁入 tiangong-ai 组织，仓库身份与历史保留。CLI 包名 @tiangong-ai/cli 和命令 tiangong-ai 不变；Wiki 新包名为 @tiangong-ai/wiki，命令不变。升级方式见迁移说明。原个人账号 tiangong-ai-legacy 保留历史；请自行 Follow 新组织。

中文 | English

TianGong AI Model Context Protocol (MCP) Server supports Streamable Http protocol.

Education graph retirement

The education knowledge graph service has been retired. The current source-built MCP server no longer advertises Search_Edu_Graph_Tool; clients must remove calls to that tool. Search_Edu_Tool continues to search education resources and does not provide graph traversal.

Previously published npm packages are not changed retroactively; npm publication is independent of this source-image deployment.

Service Deployment

GitHub CI builds reviewed main source into an ARM64 image and deploys its exact ECR digest to the existing ECS service with OIDC. See the deployment runbook for verification and rollback.

Starting MCP Server

Streamable Http Server
npm install -g @tiangong-ai/mcp-server@latest

npx dotenv -e .env -- \
npx -p @tiangong-ai/mcp-server tiangong-ai-mcp-http
Local Development Server

From this repository you can build and launch the HTTP server with:

npm run build
npx dotenv -e .env -- node dist/src/index_server.js

Launch the Inspector separately when interactive MCP inspection is needed.

Launch MCP Inspector
npx @modelcontextprotocol/inspector

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.