How I contribute to GitLab with a coding agent (and approve every push)

Description and use

An automated kit lets the author quickly open and approve merge requests in GitLab while retaining control over every push and comment.

Problem

Re‑entering GitLab contribution workflow after a long break and automating repetitive tasks while retaining final approval authority.

Outcome

The automated kit enabled the author to open 20 merge requests in one week, 15 of which were merged, across seven GitLab projects, while the author maintained final approval on every push and comment.

Workflow

• Clone the kit repository
• Run tools/init.sh with GitLab username
• Set personal access token in .env
• Verify token with python3 tools/gl.py me
• Edit AGENTS.md to specify stack and projects
• Rewrite wording-instructions.md to match personal style
• Add projects to projects.json if needed
• Run the scout workflow to rank free issues
• Run analyze workflow to generate plans and resolve questions
• Claim an issue with a one‑line comment, awaiting verifier approval
• Start and ship: implementers write code, mr-writer creates MR description
• Check pipelines and @gitlab-bot ready comment, awaiting author approval
• Post MR and monitor review comments

Evidence

In one hackathon week, this setup helped me open 20 MRs in 7 GitLab projects. 15 of them are merged now.

Source

The loop is simple: find a free issue, analyze it, claim it, implement it, ship the MR, and follow up on reviews.

Source

The claim is one line: "Hello! I'd like to work on this." The verifier checked it first, and the agent waited for my OK before posting.

Source

As of October 20, 2026, 15 of the 20 hackathon MRs are merged, 5 are in review, and none were closed.

Source