What is OmniForge and how does it automate GitLab MRs?

OmniForge is an AI-powered plugin for Claude Code that automates the entire GitLab merge request (MR) lifecycle. It uses multi-agent AI for adversarial code review (OmniReview), automatically applies and verifies code fixes (OmniFix), and intelligently creates MRs with auto-populated metadata (OmniCreate).

How does OmniForge's code review (OmniReview) work?

OmniReview employs three specialized AI agents (MR Analyst, Codebase Reviewer, Security Reviewer) that independently analyze your merge request in isolated Git worktrees. Their findings are then cross-correlated, confidence-scored, and filtered to provide a single, comprehensive review report, focusing on process quality, code correctness, and security vulnerabilities.

What is OmniFix and how does it handle code issues?

OmniFix is a skill within OmniForge designed to automate the resolution of code findings. It triages identified issues, applies the necessary fixes directly, verifies the changes, commits them, and can even resolve corresponding discussion threads in your GitLab merge request.

Can OmniForge help with creating new merge requests?

Yes, OmniCreate is the skill responsible for intelligent merge request creation. It streamlines the process by automatically generating an MR with a descriptive title and comprehensive description, derived from your commit history, saving developers significant time and effort.

What are the core technologies OmniForge utilizes?

OmniForge is built as a Claude Code plugin, leveraging multi-agent AI for its intelligence. It uses Git worktrees for isolated code analysis, interacts with GitLab via the glab CLI, and manages dependencies efficiently with 'uv' for its MCP tool server.

OmniForge

Name: OmniForge
Author: nexiouscaliver

•

Automates the GitLab merge request lifecycle through multi-agent AI for adversarial code review, automated fix application, and intelligent MR creation.

OmniForge

•

Automates the GitLab merge request lifecycle through multi-agent AI for adversarial code review, automated fix application, and intelligent MR creation.

The merge request lifecycle often suffers from inconsistent reviews, tedious manual finding resolution, and a chore-like process for creating descriptive MRs. OmniForge automates all three stages as a unified Claude Code plugin. It leverages multi-agent adversarial AI for comprehensive code review in isolated worktrees, provides automated fixing of identified findings with verification, and streamlines merge request creation by auto-populating titles and descriptions from commit history. This integrated approach aims to reduce friction and improve the quality of the entire MR workflow.

Key Features

01Multi-agent adversarial MR review with confidence scoring

02Automated finding triage, application, and verification (OmniFix)

03Intelligent merge request creation with auto-populated metadata (OmniCreate)

04Isolated code analysis using Git worktrees for each agent

05Cross-correlation and false positive reduction for review findings

063 GitHub stars

Use Cases

01Ensuring consistent and thorough code reviews for GitLab merge requests

02Automating the remediation of identified code review findings

03Streamlining the creation of well-described GitLab merge requests