GitHub Action for CLI

This action exposes the AsyncAPI CLI. It allows you to generate documentation, validate AsyncAPI documents, convert between different AsyncAPI versions and much more. The source code of the action can be found here

Inputs

cli_version

Version of the AsyncAPI CLI you wish to use. You can find all available versions here. Recommended leave it out of the inputs and use the default value.

Default points to thelatest version.

[!TIP] We recommend to default to latest version. This way there is no overhead with the script updating the CLI version. As it takes a lot of time to update the CLI version, we recommend to update it only when you need to use another one for compatibility reasons.

command

Command that you wish to run. You can find all available commands Available commands are:

• generate - generates documentation from AsyncAPI document
• validate - validates AsyncAPI document
• optimize - optimizes AsyncAPI document
• convert - converts AsyncAPI document to another version
• custom - allows you to run any command that is available in the AsyncAPI CLI. You can find all available commands here.

Default points to generate command.

[!IMPORTANT] In case you want to use custom command, you need to pass an array of commands to the custom_command input. Although passing command is not required, it is recommended to pass it to avoid any issues later on. For example, if you want to run asyncapi bundle ./asyncapi.yaml --output final-asyncapi.yaml you need to pass "bundle ./asyncapi.yaml --output final-asyncapi.yaml" to the custom_command` input.

custom_command

In case you want to use custom command you need to pass the command that you want to run in this input. You can find all available commands here.

Default points to '' (empty string).

Sample usage:

1- name: Generating HTML from my AsyncAPI document
2  uses: asyncapi/cli@v2.16.0# You can use any version you want
3  with:
4    custom_command: bundle ./asyncapi.yaml --output final-asyncapi.yaml

[!CAUTION] You have to pass the whole command as a string including the parameters and the command itself. It will run like this: asyncapi <custom_command>

filepath

Path to the AsyncAPI document that you want to process.

Default expects the AsyncAPI document to be in the root of the repository and named asyncapi.yaml.

template

Template for the generator. Official templates are listed here https://github.com/asyncapi/generator#list-of-official-generator-templates. You can pass template as npm package, url to git repository, link to tar file or local template.

Default points to @asyncapi/markdown-template@0.10.0 template.

[!TIP] We recommend to always specify the version of the template to not encounter any issues with the action in case of release of the template that is not compatible with given version of the generator.

language

Specifies the language to be used for the generated models. The value must be a valid language name supported by modelina.

Default is not set.

[!WARNING] Either language or template must be set else an error will be thrown. The action will return an error if the language is not supported by modelina.

output

Path to the output directory. Can be used for generate and convert commands.

Default points to output directory in the root of the repository.

parameters

The command that you use might support and even require specific parameters to be passed to the CLI for the generation. You can find all available parameters here.

Default points to '' (empty string).

[!NOTE] For template parameters, you need to pass them as -p <template_parameters> as can be seen in CLI documentation.

Example usage

[!WARNING] Using docker://asyncapi/github-action-for-cli will not work as expected. This is because the GitHub Actions runner does not pass params to the docker image correctly. This is why we recommend to use asyncapi/cli instead. However, you don't need to worry as it won't build the image every time. It will pull it from Docker Hub as it is already built there.

Basic

In case all defaults are fine for you, just add such step:

1- name: Generating Markdown from my AsyncAPI document
2  uses: asyncapi/cli@v2.16.0 # You can use any version you want

Using all possible inputs

In case you do not want to use defaults, you for example want to use different template:

1- name: Generating HTML from my AsyncAPI document
2  uses: asyncapi/cli@v2.16.0 # You can use any version you want
3  with:
4    command: generate
5    filepath: ./docs/api/asyncapi.yaml
6    template: "@asyncapi/html-template@0.9.0" #In case of template from npm. Or can use a link.
7    output: ./generated-html
8    parameters: "-p baseHref=/test-experiment/ sidebarOrganization=byTags"

[!IMPORTANT] Note the usage of -p in parameters input. This is required for template parameters, unlike previous versions of this action as the action includes other commands than just generate.

Example workflow with publishing generated HTML to GitHub Pages

In case you want to validate your asyncapi file first, and also send generated HTML to GitHub Pages this is how full workflow could look like:

1name: AsyncAPI documents processing
2
3on:
4  push:
5    branches: [ master ]
6
7jobs:
8  generate:
9    runs-on: ubuntu-latest
10    steps:
11    #"standard step" where repo needs to be checked-out first
12    - name: Checkout repo
13      uses: actions/checkout@v2
14      
15    #In case you do not want to use defaults, you for example want to use different template
16    - name: Generating HTML from my AsyncAPI document
17      uses: asyncapi/cli@v2.16.0 # You can use any version you want
18      with:
19        template: '@asyncapi/html-template@0.9.0'  #In case of template from npm, because of @ it must be in quotes
20        filepath: docs/api/my-asyncapi.yml
21        parameters: -p baseHref=/test-experiment/ sidebarOrganization=byTags #space separated list of key/values
22        output: generated-html
23      
24    #Using another action that takes generated HTML and pushes it to GH Pages
25    - name: Deploy GH page
26      uses: JamesIves/github-pages-deploy-action@3.4.2
27      with:
28        ACCESS_TOKEN: ${{ secrets.GITHUB_TOKEN }}
29        BRANCH: gh-pages
30        FOLDER: generated-html

Example workflow for generating models

In case you want to use models generated from your AsyncAPI document, you can use this action to generate them and then use them in your workflow. This is how full workflow could look like:

1
2name: AsyncAPI documents processing
3
4on:
5  push:
6    branches: [ master ]
7
8jobs:
9  generate-models:
10    runs-on: ubuntu-latest
11    steps:
12    #"standard step" where repo needs to be checked-out first
13    - name: Checkout repo
14      uses: actions/checkout@v2
15      
16    - name: Generating models from my AsyncAPI document
17      uses: asyncapi/cli@v2.16.0 # You can use any version you want
18      with:
19        command: generate
20        filepath: docs/api/my-asyncapi.yml
21        language: typescript
22        output: generated-models

Example workflow for validating AsyncAPI document changes

In case you want to validate your AsyncAPI document changes, you can use this action to validate them and then use them in your workflow. This is how full workflow could look like:

1name: Validate AsyncAPI document
2
3on:
4  pull_request:
5    branches: [ master ]
6
7jobs:
8  validate:
9    runs-on: ubuntu-latest
10    steps:
11    #"standard step" where repo needs to be checked-out first
12    - name: Checkout repo
13      uses: actions/checkout@v2
14      
15    - name: Validating AsyncAPI document
16      uses: asyncapi/cli@v2.16.0 # You can use any version you want
17      with:
18        command: validate
19        filepath: docs/api/my-asyncapi.yml

Local dry run

Use following commands to run and test github action locally:

1. Build docker image of github action for cli

  npm run action:docker:build

2. Execute docker image with proper arguments

  docker run -e GITHUB_WORKSPACE="" --workdir /action  -v "/home/{user}/path/to/repo":"/action" asyncapi/github-action-for-cli  "" "generate" "github-action/test/asyncapi.yml" "@asyncapi/markdown-template@0.10.0" "" "output" "" ""

Make sure to change the path of the repo and user in the command.

Troubleshooting

You can enable more log information in GitHub Action by adding ACTIONS_STEP_DEBUG secret to repository where you want to use this action. Set the value of this secret to true and you''ll notice more debug logs from this action.