🚀 init-antigravity-workflow-context-engine

Scaffold an AI Workflow for your project — with Jira, Confluence & AI Context Engine integration.

Supports Multi-Repo: one shared knowledge base, each repo runs its own AI agent.

Forked from init-antigravity-workflow — adapted for companies with their own AI Context Engine.

Quick Start

npx init-antigravity-workflow-context-engine

The CLI will guide you through the setup. Just answer the prompts.

What Changed from the Original?

| Aspect | Original (NotebookLM) | This Fork (AI Context Engine) | |--------|----------------------|-------------------------------| | Knowledge Source | NotebookLM (upload markdown files) | AI Context Engine (auto-indexes Git + Confluence) | | Code Indexing | Manual | Auto-pull from Git → local DB | | Docs Indexing | Upload to NotebookLM | Auto-extract Confluence pages → local DB | | Knowledge Writeback | Upload to NotebookLM | Create Confluence page (auto-indexed) | | Context Isolation | File prefix in NotebookLM | Confluence space IDs + Git repo scoping |

What stays the same: 4-Round Wizard, Context Manifest, Jira MCP, /pause /resume /handover, Global/Module modes.

Architecture Overview

Who Are You?

This tool supports two roles. Pick the one that matches your situation:

🏗️ I'm setting up the Main Project

You are the Lead / Architect. You define system-wide rules that all teams must follow.

When to pick this: You're the one creating the central repository or the master standards for the project.

cd ~/projects/my-main-project
npx init-antigravity-workflow-context-engine
# Select: Global (Master Architect)

What you get:

my-main-project/
├── .aiignore
├── .agentrules
└── .antigravity/
    ├── 00_MYAPP_Agent_Workflow.md    # "Constitution" — 4-Round Wizard
    └── 00_Core_Routing.md            # Full access to all knowledge sources

Your responsibilities:

• Create [Global-Convention] pages in Confluence (architecture rules, coding standards, git flow)
• The AI Context Engine auto-indexes them into the shared knowledge space
• All module teams inherit your rules automatically

💡 Important Note: After generating the Architecture Map, create supplementary convention pages in Confluence (e.g., [Global-Convention] Clean_Architecture_Rules, [Global-Convention] UI_Theme_Standards). This establishes the "Global Knowledge Pool" that all module agents rely on.

📦 I'm setting up a Sub-Module

You are a Module Owner / Team Lead. You own one specific part of the system (e.g. Payment, Auth, Booking).

cd ~/projects/payment-service
npx init-antigravity-workflow-context-engine
# Select: Module (Module Owner)
# Enter module name: Payment

What you get:

payment-service/
├── .aiignore
├── .agentrules
└── .antigravity/
    ├── 00_MYAPP_Agent_Workflow.md    # "Constitution" — scoped 4-Round Wizard
    └── 00_Core_Routing.md            # ⚠️ Restricted: Global + your module ONLY

Context Isolation — your AI agent can only access:

• ✅ [Global-Convention] pages — system-wide rules
• ✅ [Module-Payment] pages — your own module's docs
• ✅ [Module-Auth], [Module-Core] — declared dependencies (READ-ONLY)
• ❌ Other modules' pages — blocked

How dependencies work:

• Flutter projects: auto-detected from pubspec.yaml path dependencies
• Other frameworks: enter manually during setup
• Dependencies are READ-ONLY — your agent can read their docs but cannot modify them

CLI Prompts

When you run the CLI, you'll be asked:

1.  Select your role: Global / Module
2.  [Module only] Enter module name
3.  Select framework (Flutter, Laravel, React, Node.js, Python, Other)
4.  Enter project PREFIX (e.g. LC247, MYAPP)
5.  Enter Jira project key
6.  Enter Confluence Space ID(s)          ← for Context Engine indexing
7.  Enter Git repo URL(s)                 ← for Context Engine indexing
8.  [Module only] Module dependencies     ← auto-detect or manual
9.  Select AI agent (Antigravity/Cursor/Both)
10. Overwrite existing files?

How Multi-Repo Knowledge Works

Everyone uses one shared Confluence workspace. Pages are organized by prefix:

Confluence (shared workspace)
│
├── [Global-Convention] Master_Architecture.md       ← Main Project writes
├── [Global-Convention] Clean_Architecture_Rules.md   ← Everyone must follow
├── [Global-ADR] ADR-001_State_Management.md          ← System-wide decisions
│
├── [Module-Payment] Convention_API_Design.md         ← Payment team writes
├── [Module-Payment] ADR-001_Gateway_Choice.md        ← Only Payment agent reads
│
├── [Module-Auth] Convention_JWT_Flow.md               ← Auth team writes
└── [Module-Booking] ...                               ← Booking team writes

The AI Context Engine auto-indexes all pages. Each module agent only queries pages matching its routing rules.

What the AI Agent Does (4-Round Wizard)

| Round | What Happens | |-------|-------------| | 0 | No Jira ticket? Agent interviews you and creates one automatically. | | 1 | Agent gathers context from Jira, Confluence, AI Context Engine → builds a Context Manifest. | | 2 | Agent generates an Execution Plan from the manifest → you approve it. | | 3 | Agent writes code (to real repo), runs linters, updates Jira, optionally publishes knowledge to Confluence. |

The agent also supports slash commands: /pause, /resume, /handover.

Generated Files Explained

| File | What it does | |------|-------------| | .aiignore | Like .gitignore but for AI agents — hides build artifacts, node_modules, etc. | | .agentrules | Forces the AI to read the Constitution + Routing before doing anything. | | 00_[PREFIX]_Agent_Workflow.md | The "Constitution" — 4-Round Wizard, Context Manifest, status tracking, sub-agents. | | 00_Core_Routing.md | Controls which Confluence spaces, Git repos, and module docs the AI can access. |

📖 Documentation

Detailed guides with real-world examples for every use case:

| # | Guide | Description | |---|-------|-------------| | 1 | Setup — Global (Master Architect) | Step-by-step setup for the lead architect | | 2 | Setup — Module (Module Owner) | Setup for sub-module with context isolation | | 3 | Fix a Bug (Real-World Example) | Complete bug fix walkthrough using 4-Round Wizard | | 4 | Build a Feature (Real-World Example) | New feature walkthrough with existing Jira ticket | | 5 | Pause, Resume & Handover | State management across sessions and developers |

Infographics

Requirements

• Node.js ≥ 18
• An AI agent with MCP support (Jira, Confluence)
• AI Context Engine configured to index your Git repos and Confluence spaces

Version History

| Version | Changes | |---------|--------| | 2.0.0 | Fork: Replace NotebookLM with AI Context Engine. Add Confluence Space ID / Git repo prompts. Knowledge writeback via Confluence MCP. |

Based on init-antigravity-workflow — fully rewritten for AI Context Engine.

License

MIT