mcpField
L
lakelogic-mcp

lakelogic-mcp

Local
github.com/LakeLogicDeveloper Tools0Updated 1d ago

About lakelogic-mcp

lakelogic-mcp is a local MCP server for AI assistants working with Open Lakehouse Contracts (OLC). It uses LakeLogic to validate, lint, explain, test, and run contracts, while keeping file access within its project directory and never writing to a contract target.

Key features

  1. 01Validate OLC contract syntax
  2. 02Report governance findings and fixes
  3. 03Explain rules and contract structure
  4. 04Export Open Data Contract Standard
  5. 05Generate synthetic test data
  6. 06Run contracts locally with Polars or DuckDB

Use cases

  1. 01Check whether a newly written OLC contract is valid
  2. 02Find governance issues and understand their fixes
  3. 03Test a contract with intentionally invalid data
  4. 04Review passing and quarantined row counts

MCP support

It runs as a local stdio MCP server and exposes contract validation, linting, reference, data-generation, and execution tools to MCP clients. File access stays within the project directory where it was started.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"lakelogic-lakelogic": {
"command": "uvx",
"args": [
"lakelogic-mcp"
]
}
}
}
Run commanduvx lakelogic-mcp

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypilakelogic-mcp0.1.4

Tools

7

Saved tool list · updated today

ParameterTypeRequiredDescription
pathanyNo
contract_yamlanyNo

Details

From the project README

lakelogic-mcp

mcp-name: io.github.LakeLogic/lakelogic-mcp

A local MCP server that lets AI assistants (Claude Code, Claude Desktop, Cursor and any MCP client) check the Open Lakehouse Contracts (OLC) they write, using LakeLogic itself instead of guessing.

  • Runs on your machine over stdio. No network, no account, no telemetry.
  • Never touches a platform. Two tools act locally: generate_test_data writes a data file and run_contract runs a contract on it with Polars or DuckDB. Nothing is written to a contract's target.
  • Reads and writes only inside the project folder it was started in.

Tools

ToolAnswers
validate_contractIs this a legal OLC contract? Each error's location and problem
lint_contractIs it a good one? Governance findings from lakelogic lint, each with its fix
explain_ruleWhy does a lint rule exist? What it checks, why it matters, a failing and a fixed contract
contract_referenceHow is one part of a contract written? (anatomy, quality, materialization, …)
export_odcsThe contract as an Open Data Contract Standard document
generate_test_dataSynthetic data shaped by the contract, with a share of rows that deliberately break its rules
run_contractHow many rows pass, how many are quarantined, and why (Polars or DuckDB, local only)

Install

Claude Code (plugin):

/plugin marketplace add LakeLogic/lakelogic-mcp
/plugin install lakelogic-mcp@lakelogic-mcp

or directly: claude mcp add lakelogic -- uvx lakelogic-mcp

Cursor: Add to Cursor

Any other MCP client (VS Code, Claude Desktop, ...): run uvx lakelogic-mcp as a stdio server. It is also listed in the official MCP Registry as io.github.LakeLogic/lakelogic-mcp.

All of them need uv on your PATH.

Try it locally (development checkout)

cd C:\_Personal\_SaaS\lakelogic-mcp
uv venv .venv
uv pip install --python .venv\Scripts\python.exe "mcp>=1.2" pytest -e ..\open-lakehouse-contract -e ..\lakelogic
uv pip install --python .venv\Scripts\python.exe --no-deps -e .
.venv\Scripts\python.exe -m pytest
Register it with Claude Code

From the project whose contracts you want checked:

claude mcp add lakelogic -- C:\_Personal\_SaaS\lakelogic-mcp\.venv\Scripts\lakelogic-mcp.exe

Then ask, for example:

  • "Lint the contracts in this folder and fix what you find."
  • "Why is KEY-002 flagged on silver_trips.yaml?"
  • "Write a silver contract for orders and check it before you show me."
Claude Desktop / Cursor

Add to the client's MCP settings (claude_desktop_config.json, or Cursor's mcp.json):

{
  "mcpServers": {
    "lakelogic": {
      "command": "C:\\_Personal\\_SaaS\\lakelogic-mcp\\.venv\\Scripts\\lakelogic-mcp.exe",
      "args": ["--root", "C:\\path\\to\\your\\contracts\\project"]
    }
  }
}

Once published

uvx lakelogic-mcp            # Claude Code: claude mcp add lakelogic -- uvx lakelogic-mcp

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.