Skip to content

platforms.md

Latest commit

 

History

History
110 lines (82 loc) · 3.48 KB

platforms.md

File metadata and controls

110 lines (82 loc) · 3.48 KB
Raw

Platform packages

team-ai-sync ships in the native CI/CD packaging format for each supported platform:

  • GitHub Actions for GitHub repositories.
  • GitLab CI/CD Components for GitLab projects.
  • Bitbucket Pipes for Bitbucket repositories.

Cross-platform sync is not the supported operating model. The GitHub package is documented and tested for GitHub repositories, the GitLab package for GitLab projects, and the Bitbucket package for Bitbucket repositories.

GitHub Actions

Use the stable GitHub Action tag:

- uses: paladini/team-ai-sync@v1
  with:
    github-token: ${{ secrets.TEAM_SYNC_ADMIN_PAT }}
    config-path: sync-config.json

GitHub target repositories use owner/repo values in targetRepositories.

GitLab CI/CD Component

Use the component from a GitLab source project:

include:
  - component: gitlab.com/paladini/team-ai-sync/team-ai-sync@1.0.0
    inputs:
      config-path: sync-config.json

By default the component reads the token from a CI/CD variable named GITLAB_TOKEN. To use a different variable name:

include:
  - component: gitlab.com/paladini/team-ai-sync/team-ai-sync@1.0.0
    inputs:
      config-path: sync-config.json
      token-variable-name: TEAM_SYNC_GITLAB_TOKEN

GitLab target projects may use group/project or group/subgroup/project values in targetRepositories. The component creates or updates merge requests.

To publish the component in the GitLab CI/CD Catalog, mirror this repository to GitLab, enable the CI/CD Catalog project setting, and create a SemVer tag. The tag pipeline uses the release: keyword to publish the catalog release.

Bitbucket Pipe

Use the pipe from bitbucket-pipelines.yml:

pipelines:
  default:
    - step:
        name: Sync AI Assets
        script:
          - pipe: paladini/team-ai-sync:1.0.0
            variables:
              BITBUCKET_USERNAME: $BITBUCKET_USERNAME
              BITBUCKET_TOKEN: $BITBUCKET_TOKEN
              CONFIG_PATH: 'sync-config.json'

Bitbucket target repositories use workspace/repo values in targetRepositories. The pipe creates or updates pull requests.

Define both BITBUCKET_USERNAME and BITBUCKET_TOKEN. BITBUCKET_TOKEN should be a Bitbucket API token with repository read/write and pull request access. Bitbucket app passwords are not recommended for new setups.

The pipe can be used from its public Bitbucket repository and tag. Appearance in the public Bitbucket integrations list depends on Atlassian's pipe review and curation process.

OCI image

GitLab and Bitbucket wrappers run the same versioned OCI image. The image is published to both registries:

paladini/team-ai-sync:<semver>
ghcr.io/paladini/team-ai-sync:<semver>
  • Docker Hub — published from Bitbucket Pipelines on git tags (existing flow).
  • GHCR — published from GitHub Actions (.github/workflows/docker-publish.yml) on v* tags and via workflow_dispatch. When DOCKERHUB_USERNAME / DOCKERHUB_PASSWORD secrets are set on GitHub, the same workflow also pushes to Docker Hub; otherwise Docker Hub remains Bitbucket-only.

Pin production usage to a SemVer tag. Avoid relying on mutable image tags for team-wide automation.

GHCR visibility

The first GHCR package may be created as private. To make it public:

  1. Open github.com/users/paladini/packages (or Packages on the repository page).
  2. Select team-ai-sync.
  3. Package settings → Change visibility → Public.