CLI Reference
The rubberduck-index CLI keeps a local project's semantic index in sync with the server, so your MCP-connected IDE always has fresh context to query against.
Get started
rubberduck-index initCommands
| Command | Description |
|---|---|
| rubberduck-index init | Initialize and index a project (prompts for token) |
| rubberduck-index sync | Sync changed files to server |
| rubberduck-index sync --force | Re-upload all files |
| rubberduck-index watch | Watch for changes and auto-sync |
| rubberduck-index watch -d | Watch in background (daemon mode) |
| rubberduck-index stop | Stop the background watcher |
| rubberduck-index status | Show local and server index status |
| rubberduck-index list | List all your indexed projects |
| rubberduck-index remove | Remove a project from the server |
Optional Flags for init
| Flag | Description | Default |
|---|---|---|
| --token | Bearer token (skip prompt) | Prompted |
| --server | Server URL (or RUBBERDUCK_SERVER) | semantic.rubberduck.com |
| --project | Project name | Directory name |
| --include | File patterns | **/*.py |
| --no-watch | Don't start watcher | Watcher starts |
Running watch -d once means you never have to manually sync again, every save is picked up and indexed in the background.