Substrate MCP Server

Substrate MCP Server

Substrate MCP Server

Category
访问服务器

README

Substrate MCP Server

MIT License Rust Subxt

A Model Context Protocol (MCP) server for Substrate blockchains, written in Rust. This project exposes dynamic Substrate blockchain operations (querying balances, blocks, pallets, storage, events, and more) via the MCP protocol, and is fully configurable via environment variables.

Designed to interface with the subxt crate.

✨ Features

  • Query account balances and storage dynamically
  • List pallets and their entries
  • Fetch and filter events and extrinsics
  • Submit and watch dynamic signed transactions
  • Access system and block information
  • Custom RPC calls to Substrate nodes

🚀 Potential Use Cases

  1. AI-Driven Blockchain Operations

    • Integrate with LLMs (like Cursor or Claude) to allow users to ask questions in natural language (e.g., "What was the last transfer from Alice?"), which are translated into MCP tool calls.
    • Build a chatbot that can answer questions, fetch balances, or explain on-chain activity using your MCP server as the backend.
    • Use the MCP server to provide live updates on-chain activity, such as balance changes or transaction statuses, to development tools like VSCode, Cursor, Claude Code, etc.
  2. Custom Dashboards and Monitoring

    • Create custom dashboards and monitoring systems for your Substrate blockchain
    • Display real-time data and analytics on your blockchain operations
    • Set up alerts and notifications for critical events
    • Use AI agents to detect suspicious activity by analyzing events and extrinsics in real time.

🛠️ Requirements

  • Rust
  • Access to a Substrate node endpoint (WebSocket)
  • A valid signing keypair (as hex)
  • Runtime metadata file for your target chain (see below for naming and placement)

📦 Installation

Clone the repository and build:

git clone https://github.com/ThomasMarches/substrate-mcp-rs.git
cd substrate-mcp-rs
cargo build --release

⚙️ Configuration

Create a .env file in the project root with the following variables:

# WebSocket endpoint for the Substrate node
RPC_URL=wss://your-node-url.example.com

# Signing keypair as hex (32 bytes, e.g. output of subkey inspect-key --scheme Sr25519)
SIGNING_KEYPAIR_HEX=your_signing_keypair_hex_here

Generating a Signing Keypair

You can generate a keypair and get the secret seed in hex using subkey:

subkey generate --scheme Sr25519 --output-type Json

Use the secretSeed field (strip the 0x prefix if present) for SIGNING_KEYPAIR_HEX.

Obtaining and Placing Runtime Metadata

Export the runtime metadata from your node and place it at artifacts/metadata.scale:

subxt metadata -f bytes > artifacts/metadata.scale

Important: The file must be named metadata.scale and located in the artifacts/ directory before building. The build will fail if this file is missing or misnamed.

▶️ Usage

To start the MCP server:

cargo run --release

The server will start and listen for MCP requests via stdio.

🖇️ Integrating with Cursor

To use this MCP server with Cursor, you need to add it to your Cursor MCP configuration. This allows Cursor to discover and interact with your Substrate MCP server.

  1. Build your server in release mode:

    cargo build --release
    
  2. Locate the path to the built binary (typically target/release/substrate-mcp-rs).

  3. In your project (or global) .cursor/mcp.json file, add an entry for your server. For example:

    {
      "mcpServers": {
        "substrate-mcp-rs": {
          "command": "$PROJECT_ROOT_ABSOLUTE_PATH/target/release/substrate-mcp-rs",
          "args": []
        }
      }
    }
    
    • Replace the command path with the absolute path to your built binary if it differs.
  4. Restart Cursor. It should now detect and connect to your Substrate MCP server, making its tools available for use.

For more details, see the Cursor documentation or the Model Context Protocol introduction.

🧰 Available Tools

The server exposes a set of tools for interacting with a Substrate blockchain, including:

  • query_balance: Fetch the balance of an account
  • list_pallets: List all pallets in the runtime
  • list_pallet_entries: List all storage entries for a pallet
  • dynamic_runtime_call: Execute a runtime API call
  • send_dynamic_signed_transaction: Construct, sign, and send a transaction
  • query_storage: Query storage by pallet and entry
  • get_latest_events: Get all events from the latest block
  • find_events: Find specific events by pallet and variant
  • get_latest_block: Get details about the latest block
  • get_block_by_hash: Get block details by hash
  • find_extrinsics: Find extrinsics in the latest block
  • get_system_info: Get system info via RPC
  • custom_rpc: Make a custom RPC call

See src/tooling/substrate.rs for full details and parameters.

🗂️ Project Structure

  • src/main.rs: Entry point, sets up logging and starts the MCP server
  • src/tooling/: Contains the Substrate tool implementation
  • artifacts/: Place your runtime metadata file here as metadata.scale (required before building)

📈 Next Steps and Goals

  • [ ] Add E2E tests
  • [ ] Add Unit tests
  • [ ] Add more tools

🤝 Contributing

Contributions are welcome! Please open issues or pull requests. For larger changes, open an issue first to discuss your proposal.

  • Follow Rust best practices and ensure code is documented
  • Run cargo fmt and cargo clippy before submitting
  • Add tests where possible

📄 License

MIT

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选