hyperlake

Hyperlake CLI and MCP bridge.

Authenticate against the Hyperlake control plane, discover clusters, list catalogs, run SQL queries, and expose the same API through an MCP stdio server.

Install

pip install hyperlake

Requirements

• Python >= 3.8
• Supported platforms: macOS, Linux, Windows (x64 and arm64)

Usage

# Authenticate
hyperlake auth login \
  --host https://your-subdomain.example.com \
  --email you@example.com \
  --api-key YOUR_API_KEY

# Check access
hyperlake auth whoami
hyperlake clusters list
hyperlake saaslake clusters list
hyperlake saaslake connector list --cluster CLUSTER_ID

# Run queries
hyperlake query run --cluster CLUSTER_ID --sql "SELECT 1"
hyperlake saaslake query run --cluster CLUSTER_ID --connector-name github_connector --sql "SELECT * FROM repos"

# Start MCP server
hyperlake mcp serve --stdio

How it works

This package is a lightweight wrapper that downloads the platform-specific hyperlake binary from GitHub Releases on first run. The binary is a statically compiled Go program with no additional runtime dependencies.

Other installation methods

• npm: npx @cerebrixos/hyperlake@latest version
• Homebrew: brew tap cerebrixos-org/hyperlake-cli && brew install --cask hyperlake
• Docker: docker run --rm cerebrixos/hyperlake version
• Direct download: GitHub Releases

License

Proprietary