What gets imported
| Category | Source files | Description |
|---|---|---|
| MCP Servers | .mcp.json, settings.json | Model Context Protocol server connections (stdio and HTTP transports) |
| Skills | .claude/skills/*/SKILL.md | Custom skill definitions (both project-level and global) |
| Commands | .claude/commands/*.md | Legacy command files (imported as skills) |
| Plugins | settings.json | Enabled plugins and custom marketplace sources |
| Model | settings.json | Default AI model selection |
| Permission Mode | settings.json | Agent permission behavior setting |
| Instructions | CLAUDE.md, ~/.claude/CLAUDE.md | Custom agent instructions (project and global) |
| Environment Variables | .env, .env.* files | Environment variable files |
| Shell Configuration | ~/.bashrc, ~/.zshrc | Shell startup scripts |
How it works
The import process uses a secure, token-based flow that runs through your local Claude Code agent.Generate an import token
From your workspace settings or during onboarding, Whim generates a short-lived import token (valid for 1 hour). This token authorizes the local agent to send your configuration to Whim.
Run the import agent
Copy the provided prompt and paste it into your local Claude Code terminal. The agent reads your local configuration files and identifies what can be imported.
Review each category
The import agent asks you about each category individually before importing. You choose what to include and what to skip — nothing is imported without your approval.
Configuration applied
Approved settings are saved to your workspace user configuration. You can review and adjust everything later in Workspace Settings.
Imported configuration is saved to your workspace user overrides — your personal settings for this specific workspace. It does not modify workspace-wide defaults. You can later promote settings to workspace defaults or your user defaults from the settings UI.
Progress tracking
During import, Whim tracks which categories have been completed. The UI shows a live progress checklist that updates as each category is processed, displaying:- Which steps are completed
- How many items were imported per category
- When each step finished
Duplicate configurations
If you already have settings configured in Whim, imported values use merge semantics — imported settings are applied on top of existing ones. For key-value settings like environment variables, imported values override existing values with the same key. For collections like MCP servers or skills, imported items are added alongside existing ones.Token security
- Import tokens are single-use — once the import completes, the token cannot be used again
- Tokens expire after 1 hour if not used
- Each new token invalidates any previous unused token for the same user and workspace
- Tokens are scoped to a specific workspace and user — they cannot access other workspaces