| CLI | MCP server | |
|---|---|---|
| Best for | Scripts, CI/CD, terminal workflows | AI-assisted development in coding assistants |
| How you interact | Shell commands with flags | Natural language prompts |
| Requires | Terminal access | An AI client that supports MCP (Claude Code, Cursor, etc.) |
| Deterministic? | Yes — you write exact commands | No — the agent decides which calls to make |
| Good for automation? | Yes — scriptable and pipe-friendly | No — not designed for unattended automation |
Developer tools
CLI vs MCP server
Both the CLI and the MCP server give you programmatic access to Stigg, but they’re optimized for different workflows.
Use the CLI when you want direct, predictable control: scripting, CI/CD pipelines, one-off admin tasks, or any situation where you need to know exactly what API call will be made.
Use the MCP server when you’re already working inside an AI coding assistant and want to describe what you need in plain English — for example, “show me all customers on the Pro plan who haven’t used any API calls this month.”