Getting Started¶

This guide walks you through starting MyAIAssistant using the ai_assist CLI. Complete these steps first; you can add workspace-specific steps later.

Prerequisites¶

• Python 3.11+ and uv (recommended), or Docker and Docker Compose
• For development mode: Node.js 18+ and npm

Step 1: Install the ai_assist CLI¶

From the project root, build and install the CLI (and its dependency agent_core):

cd agent_core
uv build

cd ../ai_assist_cli
uv build
uv tool install ./dist/ai_assist_cli-0.1.0-py3-none-any.whl \
  --with ../agent_core/dist/agent_core-0.1.0-py3-none-any.whl \
  --force

Verify the installation:

ai_assist --help

Step 2: Initialize a workspace¶

Create a workspace where your data, knowledge, and chat history will live:

# In the directory you want as your workspace
ai_assist init

# Or specify a path and optional name
ai_assist init /path/to/workspace
ai_assist init --name my-assistant /path/to/workspace

This creates:

• A local workspace (with .ai_assist_workspace, data/, config/, etc.)
• A global ~/.ai_assist directory for shared resources, if it does not exist yet

Step 3: Run the application¶

Start the backend and frontend from your workspace:

Production (Docker):

cd /path/to/your/workspace
ai_assist run

Development (local with hot reload):

cd /path/to/your/workspace
ai_assist run --dev

When services are ready:

Press Ctrl+C to stop all services.

Step 4: (Optional) Check workspace and global resources¶

# Workspace status
ai_assist workspace status

# List registered workspaces
ai_assist workspace list

# Global shared resources (prompts, agents, tools)
ai_assist global status
ai_assist global tree

Next steps¶

• Add more steps here as you extend the guide (e.g. indexing knowledge, configuring agents).
• See User Guide for the full workflow: organizations, projects, tasks, and knowledge base.
• For all CLI commands and options, see the ai_assist_cli/README.md file in the repository.