Getting Started

Quick start guide for ParallelCode

Installation

ParallelCode is available for Windows, macOS, and Linux.

Download

Visit our download page or directly download for your platform:

Windows: Download .exe installer
macOS: Download .dmg file
Linux: Download .AppImage or .deb package

Installation Steps

Windows:

# Run the installer
ParallelCode-Setup.exe

macOS:

# Open the DMG file and drag to Applications
open ParallelCode.dmg

Linux (AppImage):

# Make executable and run
chmod +x ParallelCode.AppImage
./ParallelCode.AppImage

Linux (Debian/Ubuntu):

# Install with dpkg
sudo dpkg -i parallelcode.deb

Initial Setup

After installation, launch ParallelCode for the first time.

1. Configure Your Projects

ParallelCode works with your existing projects. No special setup needed.

Click "Add Project" or "Open Folder"
Select your project directory
ParallelCode will recognize Git repositories automatically

2. Create Your First Workspace

A workspace is a container for one or more projects.

Single Project Workspace:

Workspace: "User Auth Feature"
  └─ Project: /path/to/my-app
     Branch: feature/user-auth

Multi-Project Workspace:

Workspace: "Full Stack Development"
  ├─ Project: /path/to/frontend
  ├─ Project: /path/to/backend
  └─ Project: /path/to/shared-types

Your First Parallel Workflow

Scenario: Developing Two Features Simultaneously

The Old Way (Sequential):

Work on Feature A → Wait for AI → Review → Commit
Switch branch
Work on Feature B → Wait for AI → Review → Commit
Total time: 30-40 minutes

The ParallelCode Way (Parallel):

Step 1: Open First Workspace

1. Launch ParallelCode
2. Create workspace: "Feature A - User Login"
3. Open project: /path/to/my-app
4. Switch to branch: feature/user-login
5. Open in Cursor (or your AI tool)

Step 2: Start First Task

In Cursor:
- Write prompt: "Implement user login with JWT authentication"
- AI starts working...

Step 3: Open Second Workspace (Don't Wait!)

1. Launch ANOTHER instance of ParallelCode
2. Create workspace: "Feature B - Payment Integration"
3. Open project: /path/to/my-app
4. Switch to branch: feature/payment
5. Open in Cursor

Step 4: Start Second Task

In Second Cursor Instance:
- Write prompt: "Integrate Stripe payment API"
- AI starts working...

Now both AIs are working simultaneously!

Step 5: Review Both Results

After 10 minutes:
- Check Workspace 1: User login complete
- Check Workspace 2: Payment integration complete
- Review both, test, and commit

Total time: ~12 minutes (vs 30-40 minutes)

Multi-Project Context Workflow

Scenario: Microservices Development

You have three related services and want AI to understand all of them.

Step 1: Create Unified Workspace

1. Launch ParallelCode
2. Create workspace: "Microservices Ecosystem"
3. Add multiple projects:
   - Add: /path/to/api-gateway
   - Add: /path/to/user-service
   - Add: /path/to/payment-service

Step 2: Open in Cursor

1. Click "Open in Cursor"
2. Cursor now sees all three projects
3. AI has full context of your architecture

Step 3: Use Context-Aware AI

Example prompts:
- "Update user-service to use the new authentication 
   middleware from api-gateway"
- "Add a new endpoint in payment-service that calls
   user-service to verify user status"

AI understands:

How services interact
Shared types and interfaces
Dependencies between projects

Working with AI Tools

Cursor

ParallelCode works seamlessly with Cursor:

Open workspace in ParallelCode
Click "Open in Cursor"
Use Cursor's AI features normally
ParallelCode manages workspace and project organization

Claude (via API or Desktop)

Open workspace in ParallelCode
Use Claude in VS Code or via Desktop app
Point Claude to your workspace directory

GitHub Copilot / Other Tools

Open workspace in ParallelCode
Open your preferred editor (VS Code, etc.)
Use AI tools as normal

Key Point: ParallelCode doesn't replace your AI tools. It manages multiple workspaces so you can run multiple AI sessions in parallel.

Best Practices

1. Branch Strategy

Create a dedicated branch for each parallel task:

git checkout -b feature/task-a
git checkout -b feature/task-b
git checkout -b feature/task-c

Open each branch in a separate workspace.

2. Independent Tasks

Choose tasks that don't conflict:

✅ Different features on different files
✅ Frontend + Backend simultaneously
✅ Feature development + Bug fixes
❌ Two tasks modifying the same file (high conflict risk)

3. Workspace Naming

Use descriptive names:

✅ "User Auth - OAuth Integration"
✅ "Payment - Stripe Setup"
❌ "Workspace 1"
❌ "Test"

4. Monitor Progress

Check each workspace periodically:

Every 5-10 minutes, review AI progress
Stop tasks that are going in wrong direction
Provide feedback to AI to correct course

5. Resource Management

Your machine has limits:

2-3 parallel workspaces: Ideal for most developers
4-5 workspaces: If you have 16GB+ RAM
More than 5: Consider closing some workspaces or upgrade hardware

Common Workflows

Development + Documentation

Workspace 1: Implement feature
Workspace 2: Write documentation
Workspace 3: Write unit tests

Frontend + Backend

Workspace 1: React frontend (branch: feature/ui)
Workspace 2: Node.js backend (branch: feature/api)

Feature + Bug Fix

Workspace 1: New feature (branch: feature/new-dashboard)
Workspace 2: Critical bug (branch: fix/memory-leak)

Exploration Mode

Workspace 1: Approach A (branch: experiment/redux)
Workspace 2: Approach B (branch: experiment/mobx)
Workspace 3: Approach C (branch: experiment/zustand)

Compare results, keep the best one

Troubleshooting

"AI is slow in multiple workspaces"

Your machine might be resource-constrained:

Close some workspaces
Increase RAM allocation
Use lighter AI models

"Git conflicts when merging branches"

This is normal with parallel development:

Review both branches before merging
Use Git's merge tools
Test merged code thoroughly

"Workspace not opening project"

Check:

Project path is correct
You have read/write permissions
Git repository is valid (if using Git features)

Next Steps

Features - Learn about all features in detail
Use Cases - See how other developers use ParallelCode
Tips & Tricks - Advanced techniques for power users

Need Help?

Documentation - Full documentation
FAQ - Frequently asked questions
Community - Join our community
Support - Get help from our team