Skip to main content

Installation

Prerequisites​

  • Python 3.11 or higher
  • uv (recommended) or pip
  • Zscaler OneAPI credentials — see Authentication
uv tool install zscaler-mcp-server

Or with pipx:

pipx install zscaler-mcp-server

Or with plain pip:

pip install zscaler-mcp-server

Verify the installation:

zscaler-mcp --version

Install from source​

git clone https://github.com/zscaler/zscaler-mcp-server.git
cd zscaler-mcp-server

# With uv (creates .venv automatically)
uv sync --all-extras
source .venv/bin/activate

# Or with pip
pip install -e .

Run with Docker​

A pre-built image is published on Docker Hub:

docker pull zscaler/zscaler-mcp-server:latest

docker run --rm \
  --env-file /path/to/.env \
  zscaler/zscaler-mcp-server:latest

For HTTP transports, expose the port:

docker run --rm -p 8000:8000 \
  --env-file /path/to/.env \
  zscaler/zscaler-mcp-server:latest \
  --transport streamable-http --host 0.0.0.0

See Docker deployment for the full reference.

Run with uvx (no install)​

uvx zscaler-mcp-server

uvx resolves and runs the server in an ephemeral environment. This is the simplest path for ad-hoc use and is the configuration most editor integrations recommend.

Next steps​

  1. Configure your credentials — set ZSCALER_CLIENT_ID, ZSCALER_CLIENT_SECRET, ZSCALER_VANITY_DOMAIN, and ZSCALER_CUSTOMER_ID
  2. Review the configuration options — services, toolsets, transport, write mode
  3. Run your first prompt — verify the server end-to-end
  4. Wire it into your editor — Claude, Cursor, Gemini CLI, VS Code, Kiro