Gitlab Pipeline Debugger

GitLab CI Debugger

This skill enables Claude to investigate GitLab CI pipeline failures by:

1. Checking the current pipeline status for a branch or a specific pipeline ID
2. Identifying failed jobs
3. Retrieving failed job logs
4. Analyzing error messages and suggesting fixes

Prerequisites

• glab CLI installed and authenticated (glab auth status to verify)
• Git repository with GitLab remote configured (for automatic repo detection)

If glab is not installed or authenticated, provide instructions for setting it up:

• Install: https://gitlab.com/gitlab-org/cli#installation
• Authenticate: glab auth login

Instructions

When the user asks to check CI status, debug pipeline failures, or view job logs, use the glab CLI commands below. All commands auto-detect the GitLab project from the git remote in the current working directory. To target a different project, add -R to any command.

Handling GitLab URLs

When the user provides a full GitLab URL, parse it to extract the project path and the pipeline or job ID, then use -R to target that project. Always use the project from the URL, ignoring the local git remote.

Before constructing any command from a URL, validate the extracted values:

• Project paths must only contain letters, digits, hyphens, underscores, dots, and forward slashes.
• Pipeline and job IDs must be purely numeric.
• Job names must only contain letters, digits, hyphens, underscores, dots, colons, slashes, and spaces.

If any value fails validation, reject it and ask the user to provide a corrected URL or ID.

• Pipeline URL format: https://gitlab.com///-/pipelines/
• Extract: -R / and -p
• Example: https://gitlab.com/my-org/my-project/-/pipelines/123456

becomes glab ci get -p 123456 -d -R my-org/my-project

• Job URL format: https://gitlab.com///-/jobs/
• Extract: -R / and ``
• Example: https://gitlab.com/my-org/my-project/-/jobs/789012

becomes glab ci trace 789012 -R my-org/my-project

Commands

1. Check Current Pipeline Status

• Run glab ci get -d to get the latest pipeline for the current branch with job details
• This shows the pipeline metadata (ID, status, ref, user) and a table of all jobs with their status,

duration, and failure reason

1. Check Specific Branch

• Use the -b flag to check the pipeline for a different branch:

glab ci get -b -d

1. Check Specific Pipeline by ID

• Use the -p flag to inspect a pipeline directly by its ID:

glab ci get -p -d

1. View Job Logs

• To retrieve logs for a specific job by its numeric ID:

glab ci trace

• To retrieve logs for a job by name within a specific pipeline:

glab ci trace -p

• To retrieve logs for a job by name on the current branch:

glab ci trace

• To retrieve logs for a job by name on a specific branch:

glab ci trace -b

1. Find Merge Request for a Branch

• If the user wants MR context (title, URL, reviewers):

glab mr list -s

1. Troubleshoot CI Failures

• First, get the pipeline status to identify failed jobs (steps 1-3 above)
• Then, retrieve the full log of the failed job (step 4)
• Analyze the log output to identify the error and suggest fixes

Examples

# Check CI pipeline status for the current branch
glab ci get -d

# Debug a failed pipeline by viewing the failed job's logs
glab ci trace 789012 -R my-org/my-project

Error Handling

• If glab is not found, instruct the user to install it
• If glab auth status reports unauthenticated, instruct the user to run glab auth login
• If a pipeline or job is not found, show the error and suggest checking the ID or branch name
• If the repo cannot be detected, suggest using -R or running from within

the correct git repository

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: opendatahub-io
• Source: opendatahub-io/ai-helpers
• License: Apache-2.0
• Homepage: https://opendatahub-io.github.io/ai-helpers/

Install and usage instructions live in the source repository linked above.