Getting Started

1. Install

Install Saga from the VS Code Marketplace, or load the .vsix locally:

code --install-extension saga-0.0.1.vsix

2. Initialize your workspace

Open the Command Palette (Cmd/Ctrl+Shift+P) and run:

Init

(shown in the palette under the Saga category). This scaffolds a .saga/ folder in your workspace root and opens the Getting Started panel — a live 3-step checklist:

1. Configure an AI provider
2. Add a context file (a product brief, PRD, or technical design)
3. Generate your first epic

Each step's checkmark reflects real state on disk, so you can close the panel and reopen it later (Getting Started command) without losing track of progress.

3. Configure an AI provider

Open Settings (Open Settings in the Command Palette) and pick a provider:

• VS Code Language Model (Copilot) — no setup beyond having Copilot enabled. This is the default.
• BYOK — Anthropic, Gemini, OpenAI, or OpenRouter. Paste your API key; it's stored in VS Code's SecretStorage, never in .saga/.
• Local — point at an Ollama or LM Studio endpoint.

See AI Providers for details on each.

4. Add context

Run Add Context File to register a product brief, PRD, or design doc (.md, .txt, .pdf, .docx, or code), or Add Inline Context to paste free-form text directly.

5. Generate your first epic

Run Generate Epics. Saga reads your context files, calls your configured AI provider, and opens a review panel where you can edit, validate, and save the generated epics.

From there, right-click an epic in the Epics & Stories sidebar and choose Generate Stories for Epic to break it down into INVEST-validated stories with Gherkin acceptance criteria.

6. Push to your tracker (optional)

If you use Jira Cloud or Azure DevOps, configure it under Settings → Tracker, then right-click an epic or story to push it. See Trackers.

What's next

• Core Concepts to understand the .saga/ folder and data model
• Generating Epics & Stories for the full generation workflow
• AI Providers to pick the right model for each task