Skip to content
Conduit

mcp server

sim2l

sim2l

denphi/sim2lcommunitylistedunreviewed
Use tool

Set up

Runs on your computer. Pick your agent and operating system.

Before you start

  • uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`)
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "sim2l": {
      "command": "uvx",
      "args": [
        "sim2l"
      ]
    }
  }
}

same format Cursor, GitHub Copilot CLI, Junie, JetBrains AI Assistant, Augment (Auggie CLI), Factory Droid, Kiro, Devin CLI +12

Values in angle brackets are placeholders. Restart Claude Desktop after editing its config.Claude Desktop docs

README

synced Oct 23, 2026

sim2l - Simulation Framework with Database-Backed Persistence

sim2l is a modular, notebook-agnostic library for defining, deploying, and executing simulations as versioned, reusable artifacts.

Features

  • Database-backed persistence: All simulations stored in SQLite with full versioning
  • Notebook-friendly authoring: Define simulations in Jupyter notebooks
  • Reusable artifacts: Deploy once, execute anywhere
  • Type-safe schemas: Comprehensive input/output validation with units support
  • Pluggable executors: Local, notebook (Papermill), HUB submit, and extensible
  • Smart caching: Automatic result caching based on inputs
  • Full provenance: Track all executions with complete metadata
  • MCP integration: Expose simulation discovery, execution, and results as MCP tools

Quick Start

Installation

pip install sim2l
sim2l init

Author a Simulation

# In Jupyter notebook
%%sim2l_inputs
voltage:
  type: Number
  units: volt
  min: 0
  max: 100

%%sim2l_outputs
current:
  type: Number
  units: ampere

# Your simulation code
current_val = voltage.magnitude / resistance

# Save outputs
from sim2l import save_outputs
save_outputs(current=current_val)

# Deploy
from sim2l import deploy_simulation
deploy_simulation(
    notebook="ohms_law.ipynb",
    name="ohms_law",
    version="1.0.0"
)

Use a Simulation

from sim2l import load_simulation

sim = load_simulation("ohms_law")
result = sim.run(voltage=5.0, resistance=100)
print(f"Current: {result.outputs.current}")

Documentation

See the docs/ directory for:

MCP Server

Install the optional dependency and start the MCP server for an MCP-compatible client:

pip install "sim2l[mcp]"
sim2l mcp serve --transport stdio

The server connects to the running cache, catalog, and results services. See the MCP integration guide for its tools, service URL configuration, authentication behavior, and deployment guidance.

License

MIT License

↑↓ navigate openesc close