Copied to clipboard

gShip AI

From ticket to production — automatically.

Connect your project. Link your board. gShip AI picks your tickets, writes the code, runs QA, and deploys — while you focus on what matters. We're open for contributions! Help us build the future of autonomous shipping.

Want to contribute? Join our developer community and help shape the future of gShip AI.

Get Started View Dev Tasks
Getting Started

Autonomous setup

Everything you need to integrate gShip into your existing development environment.

STEP 01
Git Repository
Clone the project repo to your system. Azure Repo Link →
STEP 02
CLI Tooling
Install the any CLI.
Ex: npm install -g @google/gemini-cli
STEP 03
YOLO Permissions
Run the CLI with full system context.
Ex: gemini --yolo
STEP 04
DevOps Gateway (Optional)
Enable Azure DevOps integration for automated PRs and CI/CD monitoring.
1
Install Dependencies
Clone the repository and install the backend and frontend packages.
Terminal 1
$ cd server && npm install
$ cd ../client && npm install
2
Boot the System
Open two terminals to run the Orchestrator and the Dashboard concurrently.
Terminal 1 — server:5001
$ cd server
$ npm start
# Orchestrator online
Terminal 2 — client:5173
$ cd client
$ npm start
# Dashboard online
Click here to open dashboard
📁
Project Setup
Add your local repository to gShip. Click 📁 New Project in the dashboard.
Field Description
Project Name A unique display name for your project.
Repo Root Path The absolute path to your local repository.
Azure DevOps Config (Optional) Base URL, Repository Name, and Branch.
🔑
PAT Permissions
If using Azure DevOps, your Personal Access Token (PAT) must have these specific scopes enabled:
Build (Read) Code (Read & Write) Release (Read) Pull Request (Read & Write)
Tip: You can set a global fallback PAT in your root .env file if you manage multiple projects under the same organization.
1
Create & Onboard
Create a new project. gShip automatically generates architecture and project structure documentation.
2
Submit Request
Click ➕ Create Ticket. Fill in the title, type (Bug/Feature), and detailed request.
3
Run Agent
Execute start the orchestrator agent in your CLI. Agents process: Admin → Planner → Knowledge → Developer → Tester.
4
Approve & Deploy
Once STAGED, review diffs. Click Approve and run start the orchestrator agent to trigger the PR + pipeline flow.
Approval Gate
▼
Git Push (Feature Branch)
▼
Create Azure PR (Auto-merge)
▼
Pipeline Polling (15s intervals)
▼
Ticket: DEPLOYED ✅
Guardrail Detail
Risk Gate Halts for review if package.json or .env are modified.
Timeout Pipeline polling expires after 10 minutes.
Loop Guard Maximum 10 agent steps per ticket to prevent infinite loops.
Advanced Op Command / Action
Reject & Fix Send STAGED ticket back with comments.
Dual-File Sync State kept in shared-memory/*.json and tasks/*.md.
Manual PR node server/scripts/azure-pr.js <id> <branch>
Troubleshooting
Issue Solution
Polling Timeout Verify pipeline status in Azure DevOps; check triggers.
Wrong UI Status Clear browser cache or check shared-memory/ JSON.
Agent Fails Ensure CLI is authenticated (gemini --yolo).
How it works

Full loop. Zero handoff.

Five steps from backlog to production. Every step handled by AI agents that understand your codebase.

Connect
Link your codebase and project management tool. Azure DevOps, Jira, GitHub.
Plan
Reads the ticket, understands context, and creates a technical plan.
Code
AI agents write production-grade code following your project's patterns.
Test
Automated QA. Unit tests, integration tests, code review. All covered.
Deploy
Ship to production. Automatically. Safely. Every time.
Features

Ship faster than your roadmap allows

Not another copilot. A full engineering team that ships autonomously.

Connect & Go
Works with your existing stack. No migration needed. Plug in your repo, link your board, and start shipping.
Full-Loop Automation
Not just code generation. Plan, design, code, test, deploy — the entire development lifecycle, automated end to end.
Ticket Intelligence
Reads and prioritizes your backlog. Picks the right ticket. Understands context, dependencies, and scope.
Built by Engineers
Created by a team that ships real software every day. Not a demo. Not a toy. A production-grade AI engineering platform.
Always On

Your AI engineering team, always on

The engineering velocity you need, without the overhead you don't.

Full Lifecycle
Ticket to deploy. Every step handled.
Zero Handoff
No human bottleneck. No context switching.
Ship 24/7
Your team never sleeps. Features ship around the clock.
Ready to Ship?

Stop managing tickets.
Start shipping features.

Follow the autonomous setup guide and modernize your engineering lifecycle today.

Setup Now