1
Add it to Claude Code
claude mcp add arithmuse -- python -m arithmuse.server2
Make your agent remember this setup
arithmuse's config, env vars, and the gotchas you hit — recalled in every future Claude Code, Cursor, and Codex session.
npx conare@latestFree · one command · indexes the sessions already on disk. Set up in the browser instead →
What it does
- Provides standard arithmetic operations: add, subtract, multiply, and divide
- Supports both Streamable HTTP and stdio transport protocols
- Includes error handling for division-by-zero scenarios
- Minimalist Python-based architecture
Tools 4
addAdds two numbers togethersubtractSubtracts the second number from the firstmultiplyMultiplies two numbersdivideDivides the first number by the secondEnvironment Variables
MCP_TRANSPORTSets the transport protocol, use 'stdio' for local executionDEDALUS_SERVER_NAMEOptional server name for Dedalus deploymentTry it
→What is 1234 plus 5678?
→Subtract 450 from 1200.
→Multiply 15 by 24.
→Divide 100 by 4.
Original README from akshithambekar/calculator-mcp
Arithmuse
Arithmuse is a minimal Python MCP server that provides basic calculator tools:
addsubtractmultiplydivide
Requirements
- Python 3.10+
Install
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
Run locally
python -m arithmuse.server
By default, this starts Streamable HTTP transport on 0.0.0.0:$PORT (or port 8000) with MCP endpoint at /mcp.
To run with stdio transport instead:
MCP_TRANSPORT=stdio python -m arithmuse.server
Dedalus deployment notes
- Dedalus entrypoint:
python main.py
- This server uses
dedalus_mcpand serves/mcpviaserver.serve(). - Optional env var:
DEDALUS_SERVER_NAME(default:calculator-mcp; should match your Dedalus slug)
- Keep the service stateless; no persistent filesystem assumptions are required.
- Ensure your Dedalus runtime installs project dependencies from
pyproject.toml. - Health check expectation: process starts successfully, remains running, and responds on
/mcp. - This repo keeps the
arithmusepackage at project root so imports work even if the platform does notpip install .first.
Available tools
add(a, b)->a + bsubtract(a, b)->a - bmultiply(a, b)->a * bdivide(a, b)->a / b(returns an error for divide-by-zero)
Test
pytest -q