Copilot Auto-Continue ⚡

   ___           _ _     _       _         _          ___          _   _                 
  / __\___  _ __(_) | __| |_    /_\  _   _| |_ ___   / __\___  _ _| |_(_)_ __  _   _  ___ 
 / /  / _ \| '_ \| | |/ _` __|  //_\\| | | | __/ _ \ / /  / _ \| '_ \ __| | '_ \| | | |/ _ \
/ /__| (_) | |_) | | | (_| |_  /  _  \ |_| | || (_) / /__| (_) | | | |_| | | | | |_| |  __/
\____/\___/| .__/|_|_|\__,_|_| \_/ \_/\__,_|\__\___/\____/\___/|_| |\__|_|_| |_|\__,_|\___|
           |_|                                                   |_|                         

🤖 Keep Copilot going! The ultimate GitHub Copilot enhancement suite with intelligent AI Project Manager, 🌱 Begin from Scratch AI project planning wizard, chat-mode-aware autonomous development, MCP integration, and Poe API integration for automatic asset generation.

🎯 What Is This?

Copilot Auto-Continue transforms GitHub Copilot from a simple code assistant into a fully autonomous development partner. It:

• ▶ One-click Autopilot — a single Start Autopilot button continues any project (new or half-finished) end to end
• 🔄 Auto-continues when Copilot pauses with "Shall I continue?"
• 🎯 Smart mode switching — picks Plan, Agent, Edit, or Ask mode based on context
• 🔌 MCP-aware — discovers MCP servers and tools in your workspace
• 🧠 Reads your planning docs (ROADMAP, VISION, SPEC) to guide development
• 🚀 Runs autonomous sessions that build entire projects from a vision document
• 💬 Chat participant — type @auto-continue /status in Copilot Chat
• 🔧 Git & test integration — auto-commit milestones, run tests, inject fix prompts
• 📊 Tracks progress and generates session summaries
• 🤖 Poe API integration (experimental) — auto-generates images, icons & audio assets in the background while Copilot codes
• 🌱 Begin from Scratch (Premium) — describe an idea, get a full AI-generated project blueprint (VISION, ARCHITECTURE, ROADMAP, MASTER PLAN) then launch autonomous building in one click

✨ Features at a Glance

🌟 Free vs Premium

💳 Get Premium — one-time purchase, lifetime access, all future updates included.

🚀 Quick Start

⭐ The One Button: Start Autopilot

New to the extension or jumping into a project mid-way? This is all you need.

Click ▶ Start Autopilot at the top of the Copilot Auto-Continue panel (or run Start Autopilot (Continue This Project) from the Command Palette). In a single click it will:

1. ✅ Turn on the Master Switch + Smart AI Manager
2. 🤖 Start the Project Manager
3. 📖 Read your planning docs & current progress (and offer to analyze the codebase if you have none)
4. 🧠 Generate a context-aware "next step" prompt and send it to Copilot
5. 🔄 Keep the project moving, step after step

It works whether your project is brand new or half-finished, and the same button becomes a Stop control while Autopilot is running.

💡 You'll find ▶ Start Autopilot at the top of the activity view, on the view toolbar (🚀 icon), in the Command Palette, and as the primary button on the Welcome screen.

⚠️ Don't confuse it with ▶ Start Monitor Only (advanced) — that low-level action only reacts to Copilot's own "shall I continue?" questions and does not generate prompts. For actually driving a project, use Start Autopilot.

1. Install & Prerequisites

✅ VS Code 1.96.0+
✅ Python 3.8+ (in PATH)
✅ GitHub Copilot extension
✅ Windows OS (OCR uses Windows APIs)

Install from VS Code Marketplace or:

code --install-extension copilot-auto-continue-2.1.0.vsix

2. First Launch

The extension auto-starts on VS Code launch:

1. OCR service installs dependencies automatically
2. Status bar shows 🟢 ON when ready
3. Welcome screen appears with setup guide

3. Basic Usage (Auto-Continue Only)

Just open Copilot Chat and start working. When Copilot asks:

"Shall I continue with the implementation?"

The extension automatically responds!

4. Advanced Usage (AI Project Manager)

For intelligent, context-aware development:

1. Create a .projectmanager/plans/ folder in your workspace
2. Add planning documents:
   • VISION.md - What you want to build
   • ROADMAP.md - Development phases and tasks
   • CURRENT_PHASE.md - Active work items
3. Enable Project Manager:

   { "copilotAutoContinue.projectManager.enabled": true }
   

4. Start development: Click ▶ Start Autopilot (recommended — does everything in one click) or run the Begin AI-Guided Development command

📋 Project Tracker (Free)

A live, native checklist in the activity bar that mirrors your PROGRESS.md — no premium required.

Getting Started

1. Click the Project Tracker icon in the activity bar.
2. Click Create PROGRESS.md (or add your own checkbox tasks):

   ## Milestones
   
   - [ ] **M1** Define project vision
   - [ ] **M2** Set up project structure
   - [ ] **M3** Implement core features
   

3. Tasks render grouped by heading, each showing a done/total count and overall completion. Click a task to jump to its line; use the inline check button to toggle it.

Commit-Driven Ticking

Tag a task with a milestone id (**M3**, [M3], M3:, or PREP) and the tracker will tick it automatically when a git commit closes it:

git commit -m "Add login flow — Closes M3"
# or
git commit -m "M3: done"

• Powered by the built-in Git API — no git hooks or setup required.
• Deterministic & word-boundary safe: M1 never matches M10, M8 never matches M8a.
• Toggle with the tracker.commitTicking setting (default on).

Other Actions

• Open as Checklist Preview — renders the file in VS Code's native Markdown preview (toolbar button or Command Palette).
• Auto-mark completed tasks — when the AI Project Manager finishes a task, its checkbox is ticked automatically (tracker.autoMarkTasks, exact-match only).

Settings

🤖 AI Project Manager

Transform simple "yes, continue" into intelligent guidance!

How It Works

┌─────────────────────────────────────────────────────────────┐
│  Your Planning Docs          │  Copilot Output              │
│  ├── VISION.md               │  "I've implemented auth.     │
│  ├── ROADMAP.md              │   Should I continue with     │
│  └── CURRENT_PHASE.md        │   the API or database?"      │
└─────────────────────────────────────────────────────────────┘
                            ↓
          ┌─────────────────────────────────────┐
          │  🧠 AI Project Manager              │
          │  Analyzes context, generates smart  │
          │  continuation prompt                │
          └─────────────────────────────────────┘
                            ↓
┌─────────────────────────────────────────────────────────────┐
│  "Great work on auth! Per the roadmap, API endpoints are   │
│   next. Start with the user authentication endpoint using  │
│   JWT as specified in SPEC.md"                             │
└─────────────────────────────────────────────────────────────┘

PM Commands

PM Personalities

📊 Progress Tracking

Track your project's completion status automatically.

View Progress

Run 📊 Show Project Progress to see:

• Overall completion percentage
• Tasks by status (completed, pending, blocked)
• Phase-by-phase breakdown
• Recently completed tasks

How It Works

The PM parses checkbox tasks from your planning docs:

- [x] Completed task ✓
- [ ] Pending task

📝 Session Summaries

Get comprehensive reports after development sessions.

Generate a Summary

Run 📝 Generate Session Summary to create:

• Session duration and turn count
• Accomplishments extracted from prompts
• Tasks completed during the session
• Suggested next steps from roadmap

Save Options

• 💾 Save to file → .projectmanager/sessions/
• 📋 Copy to clipboard
• 👁️ View in editor

🌱 Begin from Scratch (Premium — v3.2.4)

Describe a project idea and let the AI generate a complete, production-ready planning doc set — before a single line of code is written.

How It Works

You describe an idea (name, purpose, aesthetic, features)
         ↓
  AI generates 5 planning documents in sequence:
  VISION.md → ARCHITECTURE.md → ROADMAP.md
         → MASTER_PLAN.md (chunked by phase)
         → CURRENT_PHASE.md
         ↓
  Click ⚡ Start Autonomous Building
         ↓
  AI Project Manager consumes your new plan
  and Copilot builds the project from Phase 1

Steps

1. Open the wizard — sidebar button, Welcome panel, or Dashboard → 🌱 Begin from Scratch
2. Fill in the form:
   • Project Name, Save Location, Description, Purpose, Aesthetic, Desired Features
3. Set two sliders:
   • Creativity (1–10): 1 = established patterns · 10 = borderline unhinged
   • Project Detail (1–10): controls phase count (5 phases → up to 100 phases)
4. Click 🚀 Begin Planning — live progress bar as the AI generates all docs
5. Click ⚡ Start Autonomous Building — Copilot executes the plan start to finish

Token limit proof: MASTER_PLAN.md is generated phase-by-phase so even a 100-phase, exhaustively detailed plan is produced without ever hitting model context limits — scaling to 10,000+ lines.

What Gets Generated

All files are saved to .projectmanager/plans/ in your chosen folder.

Cross-Workspace Support

If you generate a plan for a different folder, the extension writes a .scratch-autostart.json marker. On next VS Code activation in that folder you're offered Start Autonomous Building / Open Master Plan / Not Now — no manual setup needed.

🤖 Poe API Integration (Experimental — v3.2)

Connect your Poe account to automatically generate images, icons, audio, and other assets for your project — in the background while Copilot writes the code.

How It Works

Planning docs (VISION.md, ROADMAP.md, SPEC.md)
         ↓  scanned on PM start
  Asset needs detected (logo, icon, splash, sfx...)
         ↓  fire-and-forget background jobs
  Poe bot generates asset  →  saved to .projectmanager/assets/
         ↓
  VS Code notification: "Image generated: assets/images/logo.png"

Each Copilot response is also scanned in real-time — if Copilot says "we'll need a logo for this", Poe generates one automatically.

Setup

1. Get a Poe API key at poe.com/api_key
2. Command Palette → Connect Poe API — stores key in VS Code SecretStorage
3. Enable in settings: copilot-auto-continue.poe.enabled = true
4. Set your image bot: copilot-auto-continue.poe.imageBot = "DALL-E-3" (or Flux-Pro, Imagen-4, etc.)
5. Start the AI Project Manager — assets generate automatically

Poe Commands

Poe Settings

Note: Poe is entirely optional. The extension works identically without it. A separate Poe account is required (free tier available at poe.com).

⌨️ Keyboard Shortcuts

📋 All Commands

Access via Command Palette (Ctrl+Shift+P):

Core Commands

Project Manager Commands

Session & Analysis Commands

Begin from Scratch Commands

Poe API Commands (Experimental)

⚙️ Configuration

Access: File > Preferences > Settings → search "copilot auto"

Core Settings

Project Manager Settings

Safety Settings

📂 Workspace Settings: Most settings can be customized per-project. Use the command Save Settings to This Workspace to create workspace-level overrides. Settings are stored in .vscode/settings.json and loaded automatically when you open that folder.

📁 Planning Documents Structure

your-project/
├── .projectmanager/
│   ├── plans/
│   │   ├── VISION.md         # Project goals and purpose
│   │   ├── ROADMAP.md        # Development phases
│   │   ├── SPEC.md           # Technical specifications
│   │   └── CURRENT_PHASE.md  # Active work items
│   ├── assets/               # Poe-generated assets (v3.2)
│   │   ├── images/           # Generated images/icons
│   │   ├── audios/           # Generated audio files
│   │   └── manifest.json     # Asset manifest
│   └── sessions/
│       └── session-*.md      # Session summaries
└── src/
    └── ...

Example VISION.md

# Project Vision

## Overview
A real-time multiplayer game engine with WebSocket support.

## Goals
- 60 FPS performance on mobile
- <100ms latency multiplayer
- Cross-platform support

## Key Features
- Physics engine
- Particle system
- Networked game state

🛡️ Safety Features

🐛 Troubleshooting

OCR Service Issues

1. Check "OCR Service" terminal for errors
2. Verify Python 3.8+ is in PATH
3. Run: Restart OCR Service

Patterns Not Matching

1. Enable debug logging in settings
2. Check Show Log for detection attempts
3. Verify Copilot chat panel is visible

PM Not Generating Smart Responses

1. Verify projectManager.enabled is true
2. Check planning docs exist in configured folder
3. Ensure language model is available
4. Run: Reload Planning Docs

📊 Status Bar

📚 Documentation

• AI Project Manager Guide
• Development Roadmap
• Changelog

🤝 Contributing

1. Fork the repository
2. Create feature branch (git checkout -b feature/amazing)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing)
5. Open Pull Request

📄 License

MIT License - see LICENSE