MOEX MCP Server 2026
moex_mcp_server_2026
Python MCP server for Moscow Exchange quotes: share price, history, security info, top gainers and losers, dividends and portfolio value
- 0 ★ on the repository
- Last commit
Extended access
Why this level
- The server goes to the exchange API over the network on agent requests
- Prices and dividends without caveats are easily taken for advice, so they need a critical eye
Install
- Works in
- Services
Manual install
python -m pip install -e .Install into a Python 3.11 or newer virtual environment, run through `scripts/run_server.sh`.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The server goes to the exchange API over the network on agent requests
- Prices and dividends without caveats are easily taken for advice, so they need a critical eye
This is third-party code. Review the repository files before installing.
What it does
The server on the official Python MCP SDK gives an agent nine tools on Moscow Exchange ISS API data. `get_stock_price` returns the current price of a ticker such as SBER, `get_stock_history` gives trading days with open, high, low and close prices and volume for a period, and `get_stock_info` shows the ticker, name, trading board and status. Market tools: `get_market_summary` with the number of securities and leaders, `get_top_gainers` and `get_top_losers` with a `limit` parameter. Additional ones: `get_dividends` for a ticker, `compare_stocks` compares prices and changes of several shares and `calculate_portfolio_value` computes portfolio value from positions with a ticker and quantity. Transport is stdio only for now. The repository has a minimal MCP client for protocol checks: it starts the server, finds the tools and calls `get_stock_price`. The idea and approach come from the RedMadRobot custom-mcp-article teaching project. Architecture and operations docs live in the `docs` folder.
Who it is for. Those who want a simple teaching server of Moscow Exchange quotes for an agent and are ready to run it locally over stdio with no authentication or keys.
Good fit when
- An agent answers "How much is a Sberbank share today" with a live request to the exchange
- A ticker needs history with open, high, low, close and volume for chosen dates
- Gainers and losers of the current session are listed with a set limit
- The value of a portfolio of several positions is computed in one call
Not a fit when
- Bonds, funds and indices with special handling are needed: the tools are built around shares
- The server is exposed to a network or used in production: the author left security and deployment out of scope
- You need delay notes and price units with caveats on splits and coupons: the server adds no such notes
Example request
How much does a Sberbank share cost today and what did it do over the last week?MCP
- Transport
- stdio
- Authentication
- not required
Limitations
Python 3.11 or newer is required, installed with `python -m pip install -e .` in a virtual environment. Transport is stdio only, and the author left security and production deployment out of the current scope. Requests go to the public ISS API, so network access is needed and quote delay depends on exchange terms for free data.
A check without network boils down to `python -m compileall -q src tests`. The repository states no license and the code follows a RedMadRobot teaching project. The README has no ready configuration entry for Claude Code or Cursor, and running is described through the projects minimal client.
How to disable. Stop the server process and delete its entry from the MCP client configuration if you added one.
FAQ
How do I install MOEX MCP Server?
Create a Python 3.11 or newer virtual environment and run `python -m pip install -e .` in the repository root.
How do I check the server works?
Run `PYTHONPATH=src python -m clients.mcp_client`: the client starts the server over stdio, finds the tools and calls `get_stock_price`.
Which tools does the server have?
`get_stock_price`, `get_stock_history`, `get_stock_info`, `get_market_summary`, `get_top_gainers`, `get_top_losers`, `get_dividends`, `compare_stocks` and `calculate_portfolio_value`.
Is an exchange API key needed?
The README mentions no keys: the server calls the public Moscow Exchange ISS API.
How do I compute portfolio value?
Call `calculate_portfolio_value` with positions: each holds `ticker` and `quantity`.
Discussion
No questions or reviews yet.
Related
All Moscow Exchange toolsmcp-moex (AlekseyOsikov)
mcp-moex
MCP server over the Moscow Exchange API: security search, prices, history, metrics, comparison with an index and a schedule mode, no keys needed
A read-only Python MCP server for the official Moscow Exchange ISS API: stocks, bonds, FORTS futures and dividends
MOEX MCP Server with 20 tools
MOEX MCP Server
A TypeScript Moscow Exchange ISS MCP server on a vertical-slice architecture: 20 tools from quotes to a bond yield curve
MoexMcp on .NET
MoexMcp
A .NET MOEX MCP server with clean architecture: stocks, bonds, currencies, metals, comparison and performance ranking
MOEX and T-Bank Invest market data MCP
market-mcp
A read-only MOEX ISS and T-Bank Invest market data MCP server: quotes, candles, a screener, and optional portfolio access
MOEX MCP server with caching and retries
MOEX MCP сервер
An npm-published Moscow Exchange ISS MCP server with 10 tools, SQLite caching and resilient request retries
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| MOEX MCP Server 2026 (this card) | no | not required | 0 | |
| Moscow Exchange data MCP server | no | key is optional | 4 | another Moscow Exchange server, this one has nine tools with dividends and portfolio |
| MOEX MCP Server on Go | no | not required | 0 | a server without portfolio value and top gainers and losers |
README in short
The English README with a Russian version describes features, the host, stdio, server, ISS API chain architecture, tools with parameters, stack, installation, running the minimal client, checking and the source of the idea.