mcp-market-ru
MCP server for the Russian construction market — 3,395 contractor companies and 13,436 house-building projects across 18 regions. 21 tools for search, comparison, cost analytics, contractor recommendations, and quote requests.
README
MCP Market Russia
Hosted Model Context Protocol (MCP) server providing AI agents with structured access to the Russian construction market: 3,395 contractor companies, 13,436 house-building projects, across 18 regions, exposed through 24 specialized tools.
- Live endpoint: https://mcp-market.ru/mcp/
- Interactive demo: https://mcp-market.ru/demo
- Dashboard: https://mcp-market.ru/dashboard
- Protocol: MCP v2024-11-05 (Streamable HTTP transport)
Why this server
Russia's construction market is fragmented across regional registries, vendor websites, and aggregators. AI assistants that try to answer "какие каркасные дома под ключ в Подмосковье до 3 млн руб?" end up hallucinating companies or linking to dead pages. mcp-market-ru exposes curated, structured data so agents can:
- recommend vetted contractors by category / region / budget / rating,
- estimate project cost from real market prices,
- send qualified quote requests straight to a vendor (Telegram + email notification),
- compare regions, prices, and companies analytically.
Quick start
Claude Desktop
Add the server to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"mcp-market-ru": {
"command": "npx",
"args": ["mcp-remote", "https://mcp-market.ru/mcp/"]
}
}
}
Restart Claude Desktop. The 24 tools appear under the MCP icon.
Cursor, Windsurf, any Streamable-HTTP client
Point the client directly at https://mcp-market.ru/mcp/ — no wrapper required.
Python (mcp client)
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client
async with streamablehttp_client("https://mcp-market.ru/mcp/") as (read, write, _):
async with ClientSession(read, write) as session:
await session.initialize()
result = await session.call_tool("smart_match", {"brief": "каркасный дом 180 кв.м в Подмосковье до 15 млн"})
print(result)
Tools (24)
Search & catalog
| Tool | Purpose |
|---|---|
search_companies |
Search 3 395 contractors by category / region / budget / free-text query |
search_projects |
Search 13 436 house projects by area, floors, material, price |
get_company |
Full company profile: contacts, prices, rating, reviews, projects |
get_project |
Project specs, pricing, features, company contacts |
get_categories |
All categories + company count per category |
get_regions |
All 18 regions + company count per region |
get_stats |
Catalog totals + daily agent-query / lead counters |
export_search_csv |
Export companies/projects search results as CSV (UTF-8 with BOM, Excel-friendly) |
Analytics
| Tool | Purpose |
|---|---|
market_analytics |
Aggregate market view: segments, pricing bands, regional density |
market_report |
Narrative market report for a specified region |
trend_analyzer |
Company growth, price dynamics, rating changes over time |
price_comparison |
Prices across regions and categories side-by-side |
region_comparison |
Compare multiple regions on companies / projects / prices |
compare_companies |
Head-to-head comparison of 2–3 companies |
Recommendation & estimation
| Tool | Purpose |
|---|---|
find_best_companies |
Rank contractors matching budget, region, rating floor |
contractor_recommendation |
AI-weighted top picks with reasoning |
smart_match |
Natural-language Russian brief → parsed filters + top-N contractors |
calculate_cost |
Quick cost estimate from category + region + area |
project_estimator |
Detailed estimate with economy / standard / premium tiers |
Deep profiles
| Tool | Purpose |
|---|---|
company_portfolio |
Full portfolio: all projects, prices, reviews, contacts |
company_deep_profile |
Company profile plus reviews, categories, pricing history |
review_analysis |
Sentiment breakdown, themes, strengths & weaknesses |
Lead generation
| Tool | Purpose |
|---|---|
request_quote |
Submit a lead; vendor notified via Telegram + email |
get_lead_status |
Check status of a previously-submitted lead (new / contacted / won / lost) |
Pricing
| Tier | Limit | Tools |
|---|---|---|
| Free | 100 req/day | Search + catalog tools |
| Starter | 1 000 req/day | + Analytics |
| Pro | 10 000 req/day | + Recommendations + Deep profiles |
| Enterprise | custom | Everything + dedicated support |
Get an API key: https://mcp-market.ru/register
Data freshness
Data is refreshed weekly from public registries, vendor sites, and project portals. Coverage stats:
| Field | Fill rate |
|---|---|
| Description | 100 % |
| Rating | 94 % |
| Website | 62 % |
| Phone | 44 % |
| 30 % |
Last refresh: April 2026.
Architecture
- FastAPI + FastMCP over Streamable HTTP at
/mcp/ - PostgreSQL 16 for catalog & leads
- Content-negotiated landing at
/mcp/—text/htmlreturns docs,text/event-streamreturns the MCP transport - TLS via Let's Encrypt, hosted in Moscow datacenter
REST mirror available at /api/v1/... for non-MCP clients (see /api/endpoints).
License
MIT — see LICENSE.
Contact
- Site: https://mcp-market.ru/
- Issues: https://github.com/devids77/mcp-market-ru/issues
- Telegram: @Devavatar
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。