fiatdock-mcp

fiatdock-mcp

Non-custodial USDC <-> bank for AI agents. Free quotes; $0.05 USDC per session via x402; 1% commission itemised in every quote. Own funds only. PT + EU/EEA (not UK). 18+.

Category
Visit Server

README

fiatdock-mcp

Move value between USDC and a bank account from any AI agent — non-custodially (conversion, KYC and custody are handled by Transak, a licensed provider; FiatDock never touches funds).

npx -y fiatdock-mcp        # that's it — 4 MCP tools over stdio
# optional env: AGENT_PRIVATE_KEY=0x... auto-pays the $0.05 x402 fee for paid tools
Tool Cost What it does
get_quote free Live rate + all fees itemised (incl. the 1% service commission)
create_offramp_session $0.05 USDC via x402 Agent's USDC → owner's own bank account; returns a one-time checkoutUrl
create_onramp_session $0.05 USDC via x402 Owner's own fiat → USDC to the agent's wallet (address locked)
get_order_status free Track an order by partnerOrderId
  • Remote endpoint (no install): https://fiatdock.com/mcp — Streamable HTTP, stateless, CORS-enabled. Paid tools return the x402 402 challenge there (remote can't sign payments).
  • Official MCP Registry: com.fiatdock/fiatdock-mcp
  • No MCP? GET https://fiatdock.com/tools.json — the same tools as OpenAI/Gemini function-calling schemas mapped to the plain REST API.
  • Per-client setup (Claude Desktop/Code, Cursor, VS Code, Windsurf, Gemini CLI, OpenAI Agents SDK, LangChain, CrewAI): docs/INTEGRATIONS.md · runnable examples: docs/examples/

Compliance (binding): users must be 18+, in Portugal or Transak-supported EU/EEA countries (not available in the UK or restricted countries). Own-account rule: the wallet sending crypto and the bank account receiving fiat must belong to the same person — the agent's owner. No third-party funds, no aggregation, no person-to-person transfers. Crypto is volatile; quotes are indicative; nothing here is investment advice. Terms · Privacy · Risk warning

Environment

Variable Required Purpose
FIATDOCK_URL no (default https://fiatdock.com) FiatDock API base URL
AGENT_PRIVATE_KEY only for paid tools Agent wallet key used to auto-pay the $0.05 x402 fee. Without it, free tools still work and paid tools return the 402 challenge. Use a dedicated low-balance wallet; never your main key.

Claude Desktop / Cursor / Windsurf / Gemini CLI

All four read the same mcpServers shape (file: claude_desktop_config.json, ~/.cursor/mcp.json, ~/.codeium/windsurf/mcp_config.json, ~/.gemini/settings.json):

{
  "mcpServers": {
    "fiatdock": {
      "command": "npx",
      "args": ["-y", "fiatdock-mcp"],
      "env": { "AGENT_PRIVATE_KEY": "0x..." }
    }
  }
}

Claude Code

claude mcp add fiatdock -e AGENT_PRIVATE_KEY=0x... -- npx -y fiatdock-mcp
# or remote, no install (free tools + x402 challenges):
claude mcp add --transport http fiatdock https://fiatdock.com/mcp

VS Code (Copilot agent mode)

.vscode/mcp.json:

{
  "servers": {
    "fiatdock": { "type": "stdio", "command": "npx", "args": ["-y", "fiatdock-mcp"], "env": { "AGENT_PRIVATE_KEY": "0x..." } }
  }
}

OpenAI Agents SDK / LangChain / CrewAI

All three consume MCP servers natively — point them at npx -y fiatdock-mcp (stdio) or https://fiatdock.com/mcp (Streamable HTTP):

# OpenAI Agents SDK
from agents.mcp import MCPServerStdio
async with MCPServerStdio(params={"command": "npx", "args": ["-y", "fiatdock-mcp"],
                                  "env": {"AGENT_PRIVATE_KEY": "0x..."}}) as fiatdock: ...

# LangChain (langchain-mcp-adapters)
from langchain_mcp_adapters.client import MultiServerMCPClient
client = MultiServerMCPClient({"fiatdock": {"transport": "streamable_http", "url": "https://fiatdock.com/mcp"}})

# CrewAI (crewai-tools)
from crewai_tools import MCPServerAdapter
tools = MCPServerAdapter({"url": "https://fiatdock.com/mcp", "transport": "streamable-http"})

How a typical off-ramp flows

  1. get_quote (free) — agent checks the rate and the full fee breakdown.
  2. create_offramp_session — pays $0.05 in USDC automatically via x402, receives checkoutUrl + partnerOrderId.
  3. The agent forwards checkoutUrl to its human owner (valid ~5 minutes, single use). KYC happens once, ever, at the provider.
  4. get_order_status (or a signed callback) confirms COMPLETED.

Security

Found a vulnerability? Please report it privately to osama@fiatdock.com — see SECURITY.md. Never open a public issue for security reports.

License

MIT

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured