GitLab Issues
Tasks that create, update, and fetch GitLab issues.
GitLab
CertifiedTasks that interact with GitLab projects, commits, and issues via the API.
Provide a GitLab host name and access token with project identifiers to search commits, create or update issues, and retrieve project metadata for CI/CD or governance workflows.
GitLab
Certified
Tasks that interact with GitLab projects, commits, and issues via the API.
Provide a GitLab host name and access token with project identifiers to search commits, create or update issues, and retrieve project metadata for CI/CD or governance workflows.
How to use the GitLab plugin
Create issues and merge requests, search issues, and run and watch CI/CD pipelines in GitLab from Kestra flows.
Authentication
Set token to a GitLab personal, project, or group access token with scopes covering the operations you need. url defaults to https://gitlab.com — set it to your instance URL for self-hosted GitLab. Set projectId to the numeric project ID or URL-encoded project path. Store token in a secret and set all three on each task.
Tasks
issues.Create opens a new issue — set title and optionally issueDescription and labels.
issues.Search queries issues in a project — filter by search (free-text), state (opened, closed, or all), and labels.
mergerequests.Create opens a new merge request — set title, sourceBranch, targetBranch, and optionally mergeRequestDescription.
pipelines.Create starts a pipeline for a ref with optional variables. With wait: true it polls every pollInterval (default PT10S, minimum PT1S) until the pipeline finishes, fails on failed or canceled, and cancels the pipeline when maxWaitDuration (default PT1H) is reached or the execution is killed.
pipelines.Get, pipelines.Cancel and pipelines.Retry read, cancel or retry one pipeline by its global pipelineId.
pipelines.List lists pipelines newest first, filtered by status (for example FAILED), ref, sha, source (for example SCHEDULE) or updatedAfter, and sorted by orderBy (ID or UPDATED_AT, for example). limit (1 to 10000, default 20) caps the number read. fetchType: FETCH_ONE returns the first pipeline as row, FETCH returns them all, STORE writes them to internal storage page by page, and NONE returns only size.
Triggers
pipelines.Trigger polls a project and starts an execution when pipelines reach one of the watched statuses (default SUCCESS, FAILED, CANCELED); each pipeline fires once. The first poll only records the pipelines that already finished, without starting an execution. Later polls ask only for pipelines updated since the last one recorded and inspect up to maxPipelines (1 to 100, default 20) of them.
