Getting Started with CAT

This tutorial takes you from zero to running your first CAT workflow in about 10 minutes. By the end, you’ll understand how CAT works and be ready to capture your first Intent.

Prerequisites

• VS Code (or VS Code Insiders) with GitHub Copilot Chat extension
• Node.js 24+ and npm
• A Git repository where you want to use CAT

Step 1: Install CAT

Follow the installation instructions in the README to install CAT into your project.

After installation, your project will have a _cat/ folder containing:

_cat/
├── _config/           # Manifests and configuration
├── references/        # DLC process references
├── templates/         # Artifact templates
└── [module-folders]/  # Installed modules with agent and skill files

Step 2: Say Hello to cat-help

Start here, every time. cat-help is your compass — it reads your project state and tells you exactly what to do next.

Your first stop is always cat-help. Open GitHub Copilot Chat and type:

cat-help

CAT Help scans your project, detects what artifacts exist (or don’t), and tells you exactly what to do next. For a brand-new project, it will recommend starting with intent capture.

💡 Tip: Whenever you’re unsure what to do next, run cat-help. It’s context-aware and adjusts its recommendations based on your project’s current state.

Step 3: Meet Janus

Janus is the front-door coordinator for all of CAT. Think of Janus as a receptionist who routes you to the right specialist:

cat-master

Janus will greet you and ask what you’re working on. Based on your answer, you’re routed to the appropriate agent and workflow. You don’t need to memorize workflow names — Janus knows them all.

Step 4: Capture Your First Intent

Let’s create a simple Intent Statement. This is the starting point for any work in the DLC:

cat-intent-statement

Clara, the Product Intent Manager, will guide you through a series of questions:

1. Who is this for?
2. What problem are you solving?
3. Why now — what’s the trigger?
4. What outcome do you expect?
5. How will you measure success?

Answer conversationally — Clara will structure your answers into a proper artifact.

What You’ll Get

After the conversation, Clara produces an Intent Statement artifact saved to:

_cat/artifacts/intent/intent-statement-{your-title}.md

This artifact captures the “what and why” in a structured format that feeds into later phases.

Step 5: Understand the Collaboration Menu

After each workflow step that produces an artifact, CAT presents a collaboration menu:

• A (Advanced Elicitation) — Challenge the AI to reconsider and improve its output using techniques like Socratic questioning or red-teaming
• P (Party Mode) — Bring multiple agent perspectives into a roundtable discussion
• C (Continue) — Accept the current output and move forward

For your first time, choose C to keep things simple. You can always come back and use A or P later.

Step 6: Check Your Status

At any point, run:

cat-dlc-status

This scans your _cat/artifacts/ folder and shows you what’s been completed, what’s in progress, and what’s missing across all four DLC phases.

What’s Next?

You now know the basics of CAT. Here are your next steps:

Goal	Where to Go
Walk through a complete project end-to-end	First Project Tutorial
Decide between Quick vs Full DLC flow	Choosing Your Track
Understand the DLC lifecycle	DLC Overview
See all available workflows	Workflow Map
Learn about the 9 agent personas	Agent Reference
Adopting CAT with your whole team	Adopting CAT as a Team

---

See Also: DLC Overview · Workflow Map · Glossary