MCP Hub
Back to servers

SovietRepublic-MCP

An MCP server that exposes live game data from Workers & Resources: Soviet Republic to Claude by parsing game autosave files. It enables users to monitor population demographics, economic metrics, and trade history through natural language queries.

Updated
Mar 1, 2026

SovietRepublic-MCP

MCP server for Workers & Resources: Soviet Republic. Exposes live game data as tools for Claude Code and Claude Desktop.

What it does

Reads the game's autosave (stats.ini) on every tool call and returns current data — no background process, no stale cache.

Available tools

ToolDescription
get_statsFull snapshot: population, economy, year/day
get_populationCitizen counts and demographics
get_economyAll resource prices in RUB and USD
get_citizen_statusHappiness metrics (food, water, healthcare, …) on 0–1 scale
get_historyTime series for any metric across all autosave records
get_tradeCurrent period import/export totals
get_trade_historyTrade volume over time for a specific resource
list_savesAll available save folders

Requirements

  • Python 3.10+
  • Workers & Resources: Soviet Republic installed (Steam)
  • mcp[cli]==1.3.0
pip install -r requirements.txt

Setup

The server expects the game installed at the default Steam path. The autosave is read from:

<game root>/media_soviet/save/autosave1/stats.ini

If your game is installed elsewhere, update STATS_PATH in mcp_server.py.

Claude Code (~/.claude/settings.json)

{
  "mcpServers": {
    "soviet-republic": {
      "command": "python",
      "args": ["E:/SteamLibrary/steamapps/common/SovietRepublic/soviet_dashboard/main.py"]
    }
  }
}

Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "soviet-republic": {
      "command": "C:\\Users\\<you>\\AppData\\Local\\Programs\\Python\\Python313\\python.exe",
      "args": ["E:\\SteamLibrary\\steamapps\\common\\SovietRepublic\\soviet_dashboard\\main.py"]
    }
  }
}

Restart Claude after editing the config. The server starts automatically when needed.

Usage

Once connected, ask Claude naturally:

"Wie entwickelt sich meine Bevölkerung?" "What's my current food supply satisfaction?" "Show me steel import history."

Project structure

main.py          # Entry point — asyncio.run(run_mcp())
mcp_server.py    # 8 MCP tools, reads fresh from disk per call
parser.py        # Parses stats.ini → StatRecord dataclasses
requirements.txt # mcp[cli]==1.3.0

Reviews

No reviews yet

Sign in to write a review