# Langgraph Harness

> Self-hosted AI coding agent platform for GitLab, built with LangGraph. Reviews merge requests, resolves issues autonomously, and chats with full project context

- Page: https://www.whatwemade.app/p/langgraph-harness
- What it is: Web app (also AI agent), for engineering
- Made by: [Vrajpal](https://www.whatwemade.app/u/vrajpal-jhala)
- Run it yourself: get the code and run your own copy
- Live: https://vrajpal-jhala.github.io/langgraph-harness
- Code: https://github.com/vrajpal-jhala/langgraph-harness (MIT, 33 stars)
- Languages: English
- Wants contributors
- Tags: agentic-ai, ai-agents, automation, code-review, coding-agent
- Upvotes: 2, comments: 0

## About

Reviews merge requests, resolves work items and tasks autonomously, and chats with full project context — remembering what matters so every run builds on the last.

Battle-tested in production since mid-2026 — **133 releases**, 1k+ reviews, 67% comment acceptance rate (Sept 29 2026, see snapshot). Development history → every failure and fix, documented.

[](LICENSE) [](package.json) [](https://vrajpal-jhala.github.io/langgraph-harness/)

Docs · Getting Started · Features · Screenshots · Screencasts

https://github.com/user-attachments/assets/5cb90928-054c-498e-bcfb-81d21a49f7f1

Four harnesses, one shared foundation:

Shared foundation: every drafted comment is screened before it posts; a run that misbehaves (repeats a call, loops, ends on a question, skips a check) gets caught and corrected mid-run — backed by persistent checkpoints, so a run resumes instead of restarting from scratch.

## How it was made

- Claude, 50% of the work
- Stack: TypeScript, Postgres

Shared on WhatWeMade: https://www.whatwemade.app/p/langgraph-harness