Usage
Description
The init command transforms any directory into an AI-native workspace. It runs an interactive wizard that guides you through configuration, then generates all necessary files.
Options
| Flag | Description |
|---|
--hub | Force hub mode (multi-project workspace) |
--project | Force project mode (single codebase) |
-y, --yes | Skip interactive prompts, use defaults |
--agents <list> | Comma-separated list of agents to enable |
--no-memory | Disable memory system |
--no-skills | Disable skills integration |
--template <preset> | Use an identity preset (founder, developer, team, researcher) |
Examples
Interactive (Recommended)
Walks you through mode selection, naming, identity, agent selection, memory, and prompts.
Hub Mode with Defaults
npx drevon init --hub --template founder --agents copilot,claude -y
Project Mode with Specific Agents
cd my-project
npx drevon init --project --agents copilot,cursor,claude -y
All Agents
npx drevon init --agents copilot,claude,cursor,codex,windsurf,cline,aider,continue -y
What It Creates
drevon.config.json
The single source of truth for your workspace configuration.
.drevon/ directory
Contains memory, skills, and prompts subdirectories.
Memory files
Mode-appropriate memory templates (if memory is enabled).
Prompts
Index file and optional starter workflow templates.
Skills infrastructure
Skills directory with auto-installed find-skills skill and skills-lock.json.
workspace/ directory
Created in hub mode only — your project folders go here.
Agent config files
Native config files for every enabled agent.
.gitignore updates
Appends Drevon-specific entries to .gitignore.
Re-initialization
If drevon.config.json already exists, the wizard will ask if you want to overwrite it.
Re-initializing will overwrite your existing config. Back up drevon.config.json and .drevon/memory/ if they contain important data.
