Command Line Interface (CLI)

Installation

You can install @walrusai/cli via NPM or yarn. A link to the GitHub can be found here.

npm install -g @walrusai/cli
# OR
yarn global add @walrusai/cli

Usage

Once you've installed @walrusai/cli, you're almost ready to run your first test. You will need an API key, the URL to a web application you want to test, and a list of plain English test instructions. You can get an API key by signing up for free here.

Generally, each walrus.ai test requires a url, name, and instructions.

Detailed documentation about walrus.ai requests can be found here.

Passing in-line values

The quickest way to run a test with the walrus.ai CLI is by invoking the command with the appropriate values.

FlagDescription
-aYour walrus.ai API key
-nThe name of the test
-uThe URL of the application to be tested
-iThe plain English instructions
-rThe revision tag

A sample invocation looks something like this:

walrus -n 'test-name' -u https://amazon.com -a YOUR_API_KEY -r $COMMIT_ID -i \
'Login' \
'Add an item to your cart' \
'Make sure the item is in your cart'

Passing file(s)

Another method for running test(s) with the walrus.ai CLI is by defining them in YAML files, and passing these files to the CLI.

KeyDescription
nameThe name of the test
urlThe URL of the application to be tested
instructionsThe plain English instructions
variablesAny variables to be interpolated.

A sample test file looks something like this:

# test-case-1.yml
---
name: 'test-name'
url: 'https://amazon.com'
instructions:
- 'Login with :username!: and :password:'
- 'Add an item to your cart'
- 'Make sure the item is in your cart'
variables:
username!: 'walrus@walrus.ai'
password: 'hunter2'

You can then pass either a single test OR a directory of tests to the walrus.ai CLI.

walrus -a YOUR_API_KEY -f test-case-1.yml
# OR
walrus -a YOUR_API_KEY -f test-cases/

Running existing tests

Sometimes, you may want to run tests in the walrus.ai dashboard from the command line, without storing and maintaining the yml files.

You can call the walrus.ai CLI with a list of test model names:

walrus -a YOUR_API_KEY -n 'Test Name 1' 'Test Name 2'