AgentifyMe

If you are new to AgentifyMe, we recommend following the quick start guide.

The AgentifyMe CLI, agentifyme, is the primary way you'll create and deploy new workflows. The CLI also provides utility methods for developing locally, manage deployments, viewing runs and configuring secrets.

Install the CLI

You can install the CLI with our installation script:

curl -LsSf https://agentifyme.ai/install.sh | bash

Authentication

You can log in to the CLI by running airplane login. This will open a browser window to complete authentication—you can close the window when you're done. bash

agentifyme auth login

If instead you want to authenticate on behalf of a service account, perhaps to use the CLI in automation, you can set the AP_API_KEY and AP_TEAM_ID environment variables to the API key and team id of the service account:

agentifyme auth login --api-key <API_KEY>

To confirm you're logged in, you can run agentifyme auth info

agentifyme auth info
  Signed in as youremail@example.com

Updating the CLI

To update the CLI, run the following:

agentifyme update

Running BotifyMe in the local environment

AgentifyMe provides a rich development experience that allows you to quickly iterate on Workflows and Tasks, even before they've been deployed. To manage code for workflows and tasks, we recommend using the AgentifyMe CLI to start the local development server.

Required Software

AgentifyMe relies on Docker containers to manage local development environments. Ensure you install and test Docker on your machine as a first step:

Local Development Server

To launch the BotifyMe local environment, execute:

botifyme dev

This command reads the configuration file, builds sandbox environments, and constructs the required Docker images. Note that downloading Docker images depends on your internet bandwidth and may take longer during the initial setup.

You can also use the daemon mode to start and stop the server.

# Start the server
agentifyme dev up

# Stop the server
agentifyme dev down