mcp-cbr-rates
Central Bank of Russia (CBR) data for AI agents — daily and historical currency rates, key rate, inflation, and macro statistics. Five typed MCP tools, in-memory TTL cache, MIT-licensed, no API key required.
README
mcp-cbr-rates
A Model Context Protocol (MCP) server that exposes public Bank of Russia (Центральный банк РФ, CBR) data — currency quotes, key rate, inflation and a compact macro snapshot — to AI agents.
<a href="https://glama.ai/mcp/servers/atomno-labs/mcp-cbr-rates"> <img width="380" height="200" src="https://glama.ai/mcp/servers/atomno-labs/mcp-cbr-rates/badge" alt="mcp-cbr-rates MCP server" /> </a>
mcp-cbr-rates is part of the atomno family of MCP
servers focused on the Russian fintech ecosystem. It is fully open-source,
requires no API keys, and is built on top of the official public CBR
endpoints.
Features
- Five high-quality MCP tools, each with a strict Pydantic schema:
get_rate,history_rates,key_rate,inflation,statistics. - Built-in TTL (Time-To-Live) cache: 1 hour for daily quotes, 24 hours for historical series, to be polite to the source.
- Async
httpxtransport with automatic retries on 5xx errors. - Safe XML parsing via
defusedxml. - 50+ unit tests with
respx-mocked HTTP, ≥80 % coverage. - No secrets, no telemetry, no third-party trackers.
Quick start
Install from PyPI (recommended)
pipx install atomno-mcp-cbr-rates
mcp-cbr-rates # starts the MCP server over stdio
Or with uv:
uv tool install atomno-mcp-cbr-rates
Install from source
git clone https://github.com/atomno-labs/mcp-cbr-rates.git
cd mcp-cbr-rates
pip install -e .
mcp-cbr-rates # starts the MCP server over stdio
Use with Cursor
Add the following to .cursor/mcp.json (or your global ~/.cursor/mcp.json):
{
"mcpServers": {
"cbr-rates": {
"command": "mcp-cbr-rates"
}
}
}
Use with Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"cbr-rates": {
"command": "mcp-cbr-rates"
}
}
}
On Windows the config lives at
%APPDATA%\Claude\claude_desktop_config.json; on macOS at
~/Library/Application Support/Claude/claude_desktop_config.json.
Use with Claude Code
claude mcp add cbr-rates -- mcp-cbr-rates
Tools
| Name | Inputs | Returns |
|---|---|---|
get_rate |
char_code: str, on_date?: date |
CurrencyRate — single quote on the given (or latest) date |
history_rates |
char_code: str, date_from: date, date_to: date |
HistoryRates — series of daily quotes |
key_rate |
date_from?: date, date_to?: date |
KeyRateHistory — CBR key-rate series |
inflation |
year_from?: int, year_to?: int |
InflationData — monthly year-over-year CPI in percent |
statistics |
(none) | MacroSnapshot — combined dashboard: key rate + USD/EUR/CNY + inflation |
Examples in plain English:
"What was the official EUR rate on April 25, 2024?" Tool:
get_rate(char_code="EUR", on_date="2024-04-25")
"Plot the daily USD-RUB rate over the last 90 days." Tool:
history_rates(char_code="USD", date_from=..., date_to=...)
"Give me the latest key rate, USD/EUR/CNY, and inflation in one go." Tool:
statistics()
The history_rates window is capped at 366 days; for longer periods, call
the tool repeatedly.
Configuration
All settings are optional and read from environment variables:
| Variable | Default | Description |
|---|---|---|
CBR_HTTP_TIMEOUT |
15 |
HTTP timeout in seconds for CBR calls. |
CBR_CACHE_DAILY_TTL |
3600 |
Cache TTL for daily quotes (seconds). |
CBR_CACHE_HISTORY_TTL |
86400 |
Cache TTL for historical series and SOAP responses. |
CBR_LOG_LEVEL |
INFO |
Standard Python log level. |
There are no API keys to configure — all CBR endpoints used here are public.
Development
git clone https://github.com/atomno-labs/mcp-cbr-rates.git
cd mcp-cbr-rates
python -m venv .venv && source .venv/bin/activate # or .\.venv\Scripts\activate on Windows
pip install -e ".[dev]"
pytest --cov=src/mcp_cbr_rates
Layout:
apps/mcp-cbr-rates/
├── src/mcp_cbr_rates/
│ ├── server.py # FastMCP entry point, tool registration
│ ├── tools.py # high-level async tools with caching
│ ├── client.py # httpx wrapper around CBR XML / SOAP / HTML endpoints
│ ├── schemas.py # Pydantic v2 models for inputs & outputs
│ ├── cache.py # async TTL cache
│ ├── currency_codes.py # static ISO → CBR id map (with dynamic fallback)
│ └── errors.py # typed exception hierarchy
└── tests/ # respx-mocked unit tests + fixtures
Data sources
https://www.cbr.ru/scripts/XML_daily.asp— daily currency quotes.https://www.cbr.ru/scripts/XML_dynamic.asp— historical currency series.https://www.cbr.ru/scripts/XML_valFull.asp— currency code lookup.https://www.cbr.ru/DailyInfoWebServ/DailyInfo.asmx— SOAP service for the CBR key rate.https://www.cbr.ru/hd_base/infl/— monthly year-over-year inflation table.
All endpoints are read-only and free of charge.
Disclaimer
This project is not affiliated with the Bank of Russia in any way. It is an unofficial, best-effort wrapper around publicly available data. Use at your own risk; the authors disclaim any responsibility for the freshness, accuracy or applicability of the data delivered through this server.
If CBR's HTML or XML schemas change, individual tools may stop working until this package is updated. Please open an issue if you notice a regression.
License
MIT — see LICENSE.
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.