Flow MCP Server
@lmcmz
About Flow MCP Server
Model Context Protocol (MCP) server for Flow blockchain with direct RPC communication
Config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"flow-mcp-server": {
"command": "npx",
"args": [
"-y",
"@outblock/flow-mcp-server",
"--stdio"
]
}
}
}Tools
7Get Flow balance for an address
Get fungible token balance
Get account information
Resolve a .find or .fn domain to an address
Execute a Cadence script
Send a transaction to the blockchain
Get transaction details by ID
Overview
What is Flow MCP Server?
A Model Context Protocol (MCP) server that enables AI assistants to access Flow blockchain data and perform operations through a standardized interface.
How to use Flow MCP Server?
Run npx -y @outblock/flow-mcp-server --stdio for AI assistant integration, or npx -y @outblock/flow-mcp-server --port 3000 for HTTP mode. Configure via environment variables like FLOW_NETWORK and FLOW_ACCESS_NODE. Integrate with Claude by adding it to tools configuration as a command.
Key features of Flow MCP Server
- Balance checking for Flow and fungible tokens
- Domain resolution for
.findand.fndomains - Script execution for reading blockchain data
- Transaction submission and monitoring
- Account information retrieval
- Flow configuration via environment variables
Use cases of Flow MCP Server
- AI assistant checking a user’s Flow balance
- Resolving a
.findor.fndomain to an address - Executing a Cadence script to read on-chain data
- Sending a transaction and monitoring its status
FAQ from Flow MCP Server
How do I run the server?
Use npx -y @outblock/flow-mcp-server --stdio for stdio mode, or add --port 3000 for HTTP mode. You can also specify the network with --network testnet.
What networks are supported?
You can connect to mainnet, testnet, or emulator by setting the FLOW_NETWORK environment variable or using the --network flag.
How do I configure the server?
Set environment variables: PORT, FLOW_NETWORK, FLOW_ACCESS_NODE, and LOG_LEVEL. If no port is set, the server runs in stdio mode.
What tools are available?
Tools include get_balance, get_token_balance, get_account, resolve_domain, execute_script, send_transaction, and get_transaction.
How do I integrate with an AI assistant like Claude?
Add a tool configuration entry with "command": "npx -y @outblock/flow-mcp-server --stdio" to your Claude tools JSON.
Frequently asked questions
How do I run the server?
Use `npx -y @outblock/flow-mcp-server --stdio` for stdio mode, or add `--port 3000` for HTTP mode. You can also specify the network with `--network testnet`.
What networks are supported?
You can connect to mainnet, testnet, or emulator by setting the `FLOW_NETWORK` environment variable or using the `--network` flag.
How do I configure the server?
Set environment variables: `PORT`, `FLOW_NETWORK`, `FLOW_ACCESS_NODE`, and `LOG_LEVEL`. If no port is set, the server runs in stdio mode.
What tools are available?
Tools include `get_balance`, `get_token_balance`, `get_account`, `resolve_domain`, `execute_script`, `send_transaction`, and `get_transaction`.
How do I integrate with an AI assistant like Claude?
Add a tool configuration entry with `"command": "npx -y @outblock/flow-mcp-server --stdio"` to your Claude tools JSON.
Basic information
More Other MCP servers
ICSS
chokcoco不止于 CSS
MCP Go 🚀
mark3labsA Go implementation of the Model Context Protocol (MCP), enabling seamless integration between LLM applications and external data sources and tools.
🚀 Model Context Protocol (MCP) Curriculum for Beginners
microsoftThis open-source curriculum introduces the fundamentals of Model Context Protocol (MCP) through real-world, cross-language examples in .NET, Java, TypeScript, JavaScript, Rust and Python. Designed for developers, it focuses on practical techniques for building modular, scalable,
FastMCP v2 🚀
jlowin🚀 The fast, Pythonic way to build MCP servers and clients.
Peekaboo MCP – lightning-fast macOS screenshots for AI agents
steipetePeekaboo is a macOS CLI & optional MCP server that enables AI agents to capture screenshots of applications, or the entire system, with optional visual question answering through local or remote AI models.
Comments