Skip to main content
The CrewForm CLI (npx @crewformhq/cli) is a standalone command-line tool that lets you create, run, and manage AI agents locally — with optional connectivity to the CrewForm platform.

Installation

CLI 0.3.0 uses MIT and includes the offline walkthrough, diagnostics and named device profiles. The original 0.2.0 npm artifact retains its original licence.

Quick Start

Commands

Offline walkthrough and diagnostics

Use the published package directly:
Alternatively, build from a checkout with npm --prefix cli ci && npm --prefix cli run build, then run:
demo runs the real sequential pipeline with scripted outputs. It needs no account, backend, provider key, installed agent or network access. The JSON pack is marked simulated: true and humanReview: required; it is not an AI evaluation. doctor probes executable versions only. A successful check does not prove login, subscription eligibility, ACP compatibility or model availability.

Local Execution

Platform (API-Connected)

Running Agents

Basic Usage

Input & Output

Interactive Chat

Start a REPL session with conversation history:
In-chat commands:

Pipeline Teams

Run multi-agent workflows locally. Agents execute in sequence, with each step’s output passed as context to the next.

Team Config Example

Fan-Out (Parallel Execution)

Steps with type: "fan_out" run multiple agents in parallel, then merge results:
Failure modes:
  • fail_fast — Stop all branches if any fails
  • continue_on_partial — Collect results from successful branches

MCP Server Integration

Connect agents to external MCP servers for dynamic tool discovery:

MCP Server Config

Create a mcp-servers.json file:
Supported transports: stdio, sse, streamable-http

Platform Integration

Connect the CLI to your CrewForm workspace to browse, download, and dispatch agents remotely.

Authenticate

Get your API key from Settings → API Keys in the CrewForm dashboard.
Credentials are saved to ~/.crewform/config.json.

Browse Your Workspace

Download Agents Locally

Pull an agent or team from the platform and run it locally:

Dispatch Work Remotely

Send a task to a cloud-hosted agent or team without running locally:

Config Formats

The CLI accepts three config formats:
Exported via Agent → Menu → Export in the dashboard:
See the Pipeline Teams section above for the full team config format.

Built-in Tools

Plus any tools discovered from connected MCP servers.

Environment Variables

See the CLI README for the full list of supported providers and their environment variables.

CI/CD Integration

The CLI is designed for automation:
The legacy code_interpreter tool is disabled: model-generated JavaScript must not execute inside the CLI process. Use a separately isolated execution service for code tools.