Prerequisites
Before you begin, make sure you have:
- A Crewship account (sign up here)
- A CrewAI project ready to deploy (or use our example)
Install the CLI
curl -fsSL https://www.crewship.dev/install.sh | bash
Invoke-WebRequest -Uri "https://api.crewship.dev/cli/releases/latest/download?platform=windows-x64" -OutFile "$env:LOCALAPPDATA\Programs\crewship.exe"
[Environment]::SetEnvironmentVariable("Path", "$([Environment]::GetEnvironmentVariable('Path', 'User'));$env:LOCALAPPDATA\Programs", "User")
Authenticate
Login to connect the CLI to your Crewship account:
This opens a browser window to authenticate. Once complete, your credentials are stored locally.
Run crewship whoami to verify your authentication status.
Create a Crew
If you don’t have an existing CrewAI project, create one:
# Install CrewAI
pip install crewai
# Create a new crew project
crewai create crew my_crew
cd my_crew
Add Crewship Configuration
Create a crewship.toml file in your project root:
name = "my-crew"
framework = "crewai"
[build]
entrypoint = "src.my_crew.main:kickoff"
python = "3.11"
The entrypoint should point to the function that kicks off your crew. For CrewAI, this is
typically the kickoff method.
Deploy
Deploy your crew with a single command:
You’ll see output like:
📦 Packaging crew...
☁️ Uploading build context...
🔨 Building image...
✅ Deployed successfully!
Deployment: dep_abc123xyz
Production URL: https://my-crew.crewship.run
Invoke Your Crew
Run your deployed crew:
crewship invoke --input '{"topic": "AI agents"}'
Stream Events
Watch your crew execute in real-time:
crewship invoke --input '{"topic": "AI agents"}' --stream
▶ Run started: run_xyz789
├─ Agent: Researcher starting task...
├─ Agent: Researcher completed task
├─ Agent: Writer starting task...
├─ Agent: Writer completed task
├─ Artifact: report.md (2.4 KB)
✅ Run completed in 45.2s
View in Console
Open the Crewship Console to:
- View run history and logs
- Download artifacts
- Manage environment variables
- Monitor usage and costs
Next Steps
