Ship your first workflow in ten minutes
This guide walks you through installing the SaaSyTech CLI, signing in and running a workflow that posts a daily summary to your team.
Overview
A workflow is a sequence of steps that starts with a trigger, such as a schedule or an incoming event, and ends with an action. You can build workflows in the visual editor or describe them as code, and both stay in sync.
You will need Node.js 18 or later and a SaaSyTech workspace. If you do not have one yet, start a free trial.
Install the CLI
Install the package globally with your package manager of choice, then confirm that it is on your path.
- Check your Node version. Run
node --versionand make sure it prints 18 or higher. - Install the package. Use the command above. Prefer not to install globally? Run it with
npx @saasytech/cli. - Verify.
saasy --versionshould print the installed version.
Authenticate
Sign in from your terminal. A browser window opens so you can approve the session, and the CLI stores a short-lived token on your machine.
For CI pipelines, create an API key in Settings, then Developers and expose it as an environment variable.
Your first workflow
Create a file named daily-summary.yml. It runs every weekday at 9:00, asks the assistant for yesterday’s highlights and posts them to a channel.
Define the trigger
Run and test
Do a dry run first. It executes every step but skips actions that change data, so you can check the output safely.
- Review the output. Check the summary text printed by the
highlightsstep. - Publish. Run
saasy deploy daily-summaryto turn on the schedule. - Watch the first run. Open the run history in the dashboard to follow it live.
Triggers and schedules
Besides schedules, workflows can start from webhooks, incoming messages or changes in a connected app. A single workflow can have several triggers, and each run records which one fired.
Browse the integration directory to see which apps can act as triggers.
Deploy and monitor
Every deploy creates a new immutable version, so you can roll back with one command. Failed runs retry automatically with backoff, and you can subscribe a channel to failures from the dashboard.