trading212-mcp

trading212-mcp

MCP server for the Trading 212 public API, enabling account, portfolio, order, pie, and history access with read tools always available and trading tools opt-in.

Category
Visit Server

README

trading212-mcp

MCP server for the Trading 212 public API — equities and ETF investing on Invest and Stocks ISA accounts. It exposes the account, portfolio, order, pie, and history endpoints as MCP tools, defaults to the demo (paper trading) environment, and keeps trading tools hidden unless explicitly enabled.

Features

  • Read coverage of the whole public API: account summary, instrument universe, exchange schedules, open positions, pending orders, pies, historical orders, dividends, cash transactions, and CSV exports.
  • Trading tools (order placement/cancellation, pie management, export requests) are opt-in via an environment flag and stay invisible to the MCP client otherwise.
  • Demo environment by default; live is an explicit choice.
  • Responses are trimmed to the fields a trader acts on, and list tools accept a limit parameter to keep payloads small.
  • One automatic retry on HTTP 429, honouring the Retry-After header capped at 30 seconds (Trading 212 rate limits are strict and per-endpoint).
  • stdio transport by default, HTTP (/mcp) on request.

Tools

Read tools (always registered):

Tool Description
get_account_summary Account summary: cash breakdown, investments, and overall result
list_exchanges Exchanges with working schedules trimmed to session open/close events
list_instruments Search the tradable instrument universe by ticker/name substring and type
get_portfolio List all open positions with quantity, prices, and unrealized P/L
get_position Fetch a single open position by its Trading 212 ticker
list_orders List all pending equity orders
get_order Fetch one pending equity order by its id
list_pies List all pies with money-in, progress, and performance
get_pie Fetch one pie's full definition and per-instrument breakdown
list_historical_orders List executed equity orders, newest first (cursor-paginated)
list_dividends List paid-out dividends (cursor-paginated)
list_transactions List cash transactions: deposits, withdrawals, fees, transfers (cursor-paginated)
list_exports List requested CSV export reports and their processing status

Write tools (registered only when T212_MCP_ENABLE_TRADING is truthy):

Tool Description
place_market_order Place a market order (positive quantity = BUY, negative = SELL)
place_limit_order Place a limit order with a limit price and DAY/GTC validity
place_stop_order Place a stop order that triggers a market order at the stop price
place_stop_limit_order Place a stop-limit order (limit order placed once the stop triggers)
cancel_order Cancel a pending equity order by its id
create_pie Create a new pie from a ticker-to-weight allocation map
update_pie Replace an existing pie's definition (name and full allocation)
delete_pie Delete a pie by ID (irreversible)
duplicate_pie Duplicate an existing pie under a new name
request_export Request an asynchronous CSV export report of account history

Configuration

All configuration is via environment variables — never commit credentials.

Variable Default Purpose
T212_API_KEY — (required) API key (Basic auth username), generated in the Trading 212 app
T212_API_SECRET — (required) API secret (Basic auth password)
T212_ENV demo Target environment: demo (paper trading) or live (real money)
T212_MCP_ENABLE_TRADING off Set to true/1/yes/on to register the write tools

Getting started

Run straight from the repository with uv:

uvx --from git+https://github.com/florinel-chis/trading212-mcp trading212-mcp

The server speaks stdio by default; add --transport http --port 8000 to serve MCP over HTTP at /mcp instead. HTTP binds 127.0.0.1 by default, and a non-loopback --host is refused unless --allow-remote is also passed — read the Safety section before using that flag.

MCP client configuration

Add the server to your MCP client's configuration (stdio, via uvx):

{
  "mcpServers": {
    "trading212": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/florinel-chis/trading212-mcp", "trading212-mcp"],
      "env": {
        "T212_API_KEY": "your-api-key",
        "T212_API_SECRET": "your-api-secret"
      }
    }
  }
}

Or run the Docker image (build it first, see below):

{
  "mcpServers": {
    "trading212": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "T212_API_KEY",
        "-e", "T212_API_SECRET",
        "trading212-mcp"
      ],
      "env": {
        "T212_API_KEY": "your-api-key",
        "T212_API_SECRET": "your-api-secret"
      }
    }
  }
}

Docker

docker build -t trading212-mcp .

# stdio (what MCP clients spawn)
docker run -i --rm -e T212_API_KEY -e T212_API_SECRET trading212-mcp

# HTTP at http://127.0.0.1:8000/mcp — publish the port on loopback only:
# the MCP endpoint is unauthenticated (see Safety)
docker run --rm -p 127.0.0.1:8000:8000 -e T212_API_KEY -e T212_API_SECRET \
  trading212-mcp --transport http --host 0.0.0.0 --port 8000 --allow-remote

(--host 0.0.0.0 binds inside the container so the port mapping works — which is why --allow-remote is needed; the 127.0.0.1: prefix on -p keeps the endpoint reachable from this machine only.)

Safety

  • The HTTP transport has no authentication: anyone who can reach the /mcp endpoint can call every registered tool — read the account, and place/cancel orders or delete pies if T212_MCP_ENABLE_TRADING is set — all signed with your API key. Keep it bound to 127.0.0.1 (the CLI default; with Docker, publish as -p 127.0.0.1:8000:8000) or put it behind an authenticating reverse proxy. Never expose it directly on a public or shared network.
  • The server defaults to the demo (paper trading) environment; set T212_ENV=live deliberately.
  • Trading tools are not registered — invisible to the MCP client — unless T212_MCP_ENABLE_TRADING is truthy.
  • The Trading 212 API is v0 beta: order endpoints are not idempotent (resending a request may create duplicate orders), and rate limits are strict and per-endpoint (each tool documents its limit).
  • Use at your own risk. Nothing here is investment advice.

Development

uv sync
uv run pytest -q
uv run ruff check .

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
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
Qdrant Server

Qdrant Server

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

Official
Featured