Quickstart

This guide walks you through signing up, trying the AI chat, creating an organization, and connecting from the CLI.

1. Sign up

Create an account at broomva.tech. You can sign in with:

• Email + password — standard email/password registration
• Google — OAuth sign-in
• GitHub — OAuth sign-in

Accounts with the same email address are automatically linked, so you can sign in with any method and access the same workspace.

2. Try the chat

Navigate to broomva.tech/chat to start a conversation. The chat supports:

• Multiple AI models — Claude (Anthropic), GPT (OpenAI), Gemini (Google), and local models through Ollama and OpenRouter
• Rich message rendering — code blocks with syntax highlighting, LaTeX math, Mermaid diagrams, tables
• Attachments — upload images and documents directly into the conversation
• Memory vault — the platform remembers context across conversations using a Lago-backed knowledge graph
• MCP tools — connect external tools and services through the Model Context Protocol

Select a model from the dropdown, type a message, and send. Conversations are automatically saved to your account.

3. Create an organization

Organizations are the unit of billing, access control, and resource isolation on the platform. Every user gets a personal workspace by default, but for team use you should create a dedicated organization.

1. Open the Console
2. Click New Organization
3. Choose a name and slug (the slug becomes part of your API namespace)
4. Invite team members by email

Organization members share a credit pool and can access shared conversations, API keys, and deployment configurations.

4. Install the CLI

The BroomVA CLI (@broomva/cli) lets you manage prompts, skills, and agent context from your terminal.

# Install globally with npm
npm install -g @broomva/cli

# Or run directly with npx
npx @broomva/cli --help

Verify the installation:

broomva --version
# 0.1.0

5. Authenticate

The CLI uses the OAuth 2.0 Device Authorization Grant (RFC 8628) for secure, browser-based authentication:

broomva auth login

This will:

1. Request a device code from the server
2. Display a URL and a short code
3. Open the URL in your browser where you approve the device
4. Poll until authorization completes and store the token locally

The token is saved to ~/.broomva/config.json. You can also authenticate manually with a token from the web UI:

broomva auth login --manual

Verify your auth status:

broomva auth status
# Authenticated (token from device-flow)

6. Make your first API call

With your token stored, you can call the platform API directly:

# Using curl with the stored token
TOKEN=$(broomva auth token)

curl -X POST https://broomva.tech/api/chat \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-20250514",
    "messages": [
      { "role": "user", "content": "Hello from the API!" }
    ]
  }'

Or use the CLI's built-in commands:

# List available prompts from the platform
broomva prompts list

# List installed skills
broomva skills list

# Show current context
broomva context show

7. Explore the platform

What's next

• Upgrade your plan — the Pro tier gives you 5,000 credits/month and access to all models
• Deploy a managed instance — run your own Life agent runtime in the cloud
• Build with the SDK — use the TypeScript SDK to integrate BroomVA into your applications