Projects YAML Reference#

This document is the YAML grammar reference used for adding and editing Semaphore 2.0 projects via the sem command line utility.

Projects can also be added to Semaphore 2.0 with the sem init command. In such a case, you will not need to create a YAML file. However, sem init requires a local copy of the GitHub repository (even if that repository is not up to date) that does not contain a .semaphore/semaphore.yml file. Essentially, sem init is simpler but less flexible than the sem create command.

To learn more about the sem commands for project manipulation, you can visit the sem command line tool reference.

Properties#

apiVersion#

The apiVersion property defines the version of the YAML grammar that will be used to define the YAML file. Different YAML versions might have different features.

Here is a list of values for apiVersion: v1alpha.

kind#

The value of the kind property is a string that specifies the type of the YAML file, i.e. whether it is for creating new projects or new secrets. For projects, the value of the kind property should be Project.

Here is a list of values for kind: Project.

metadata#

The metadata property is for grouping other properties. Currently, the only supported property is name.

name#

The name property defines the name of a Semaphore 2.0 project as it will be displayed in the Semaphore 2.0 user interface, and is the output of the sem get project command.

The value of the name property should be unique among all Semaphore 2.0 projects belonging to the same organization and must only contain [a-z], [A-Z] or [0-9] characters, dashes, and underscores – spaces are not allowed.

Using the same YAML file with different name values will create multiple Semaphore 2.0 projects connected to the same GitHub repository.

spec#

The spec property is currently used for holding the required repository and optional schedulers properties.

repository#

The repository property is used for holding the url, run_on, and, forked_pull_requests properties.

url#

The url property is a string that specifies the URL of the GitHub repository you want to add to Semaphore 2.0. The format of the url value should be as follows:

git@github.com:github_username/github_repository.git

If the value of url is erroneous, you will get various types of error messages, as detailed below.

  1. If the GitHub repository cannot be found, sem create will reply with the following error message:
$ sem create -f goDemo.yaml

error: http status 422 with message "{"message":"repository \"text/goDemo\" not found"}" received from upstream
  1. If the Semaphore 2.0 project's name is already taken, sem will reply with the following error message:
$ sem create -f goDemo.yaml

error: http status 422 with message "{"message":"project name \"goDemo2.1\" is already taken"}" received from upstream
  1. If the url value does not have the correct format, sem will print the following error message:
$ sem create -f goDemo.yaml

error: http status 422 with message "{"message":"repository url must be an SSH url"}" received from upstream
run_on#

The value of the run_on property is an array of repository events which should trigger new workflows for your Semaphore project.

Here is a list of values for run_on: branches, tags, pull_requests, and forked_pull_requests.

For more information about workflow triggers, visit the project workflow tigger options documentation page.

forked_pull_requests#

The forked_pull_requests property is used for holding the allowed_secrets and allowed_contributors properties.

allowed_secrets#

The allowed_secrets property specifies the array of secrets' names that are allowed to be exported into jobs triggered by forked-pull-requests. If the array is empty, no secrets will be exported.

allowed_contributors#

The allowed_secrets property specifies an array of secrets (i.e. their names) that are allowed to be exported into jobs triggered by forked-pull-requests. If the array is empty, no secrets will be exported.

pipeline_file#

The pipeline_file property is used for setting the initial pipeline file that is executed when a post-commit hook is received by Semaphore.

The default value is .semaphore/semaphore.yml.

Note: When modifying the initial pipeline file it is necessary to also update the corresponding GitHub status check.

status#

The status property is used to specify which Semaphore pipeline(s) will submit a status check on GitHub pull requests.

A pipeline can create a single status check as a result of the whole pipeline. Or each block in a pipeline can create its own status check.

pipeline_files#

The pipeline_files property is a list of pipeline files for which Semaphore will submit a status check.

Each value has two properties: path and level.

When you create a project, the default value is path: .semaphore/semaphore.yml, level: pipeline.

path#

The path property specifies a pipeline.

level#

The level property specifies the granularity of status checks.

Here is a list of values for level: block, pipeline.

schedulers#

The schedulers property can contain a list of schedulers defined in the project.

A scheduler is a way to run a pre-defined pipeline on a project at the pre-defined time. All times are interpreted as UTC.

A scheduler has the following properties:

name#

The name property uniquely identifies each scheduler within an organization.

branch#

The branch property specifies which branch will be used for running the pipeline.

The chosen branch must have at least one workflow initiated by a push in order for the scheduler to be valid.

at#

The at property defines the schedule under which the pipeline will be run.

Semaphore expects this property to be in the standard cron syntax. For a simple way to define your cron syntax, visit crontab.guru.

pipeline_file#

The pipeline_file property contains the relative path to the pipeline definition file from the root of the project.

For more information on defining a valid pipeline file, visit the Pipeline YAML Reference documentation page.

status#

The current status of the scheduler. This can be ACTIVE if the scheduler is currently enabled, or INACTIVE if the scheduler is currently disabled.

Examples#

A simple project:

apiVersion: v1alpha
kind: Project
metadata:
  name: goDemo
spec:
  repository:
    url: "git@github.com:renderedtext/goDemo.git"
    run_on:
      - branches
      - tags
    pipeline_file: ".semaphore/semaphore.yml"
    status:
      pipeline_files:
        - path: ".semaphore/semaphore.yml"
          level: "pipeline"

A project with schedulers:

apiVersion: v1alpha
kind: Project
metadata:
  name: goDemo
spec:
  repository:
    url: "git@github.com:renderedtext/goDemo.git"
    run_on:
      - branches
      - tags
    pipeline_file: ".semaphore/semaphore.yml"
    status:
      pipeline_files:
        - path: ".semaphore/semaphore.yml"
          level: "pipeline"
  schedulers:
    - name: first-scheduler
      branch: master
      at: "5 4 * * *"
      pipeline_file: ".semaphore/cron.yml"
    - name: second-scheduler
      branch: master
      at: "5 3 * * *"
      pipeline_file: ".semaphore/semaphore.yml"
      status: ACTIVE

See Also#