Skip to content

Quick Start

This guide targets stable 0.6.2. You do not need an account or a model subscription to start organizing local assets.

Choose your operating system and CPU architecture on the download page: DMG for macOS, an installer for Windows, or an available Linux package. Launch the app to open your local workspace.

0.6.3-beta.1 is a preview focused on Skill upgrade startup and backup restoration. To try it, back up first and use the preview release. Stable downloads remain on 0.6.2.

  1. Open Prompts and create an entry with a title and System Prompt or User Prompt.
  2. Use {{topic}} for changing values, for example “Summarize: {{content}}”.
  3. Save and assign folders or tags as needed.
  4. Copy or test the prompt from its details, filling in the variable form first.

Use cards or a list for text, and the gallery for image prompts with covers. Configure a provider and model in AI Model Settings before running AI tests.

  1. In Skills, import from a store, local directory, or Git source.
  2. Check SKILL.md and the scripts, references, and images it uses.
  3. Select Agent or project targets and check their installation paths.
  4. Apply changes, inspect each result, and confirm the Skill appears in the target tool.

Copies are independent and need redistribution after edits. Links refer to the source directory, which must remain accessible. See Skill Management.

In the Agent Workspace, select a tool and check its installation and root directory. Inspect its Skills, MCP, Plugins, Rules, and configuration. Available operations depend on the selected platform.

Create a full backup in data settings and keep the resulting .phub.gz file. After CLI or external edits, refresh the workspace to load changes; save your current drafts first.

Set up WebDAV / S3 sync or self-hosted backups when needed. Official cloud accounts, community access, and membership are separate: signing in does not upload your entire local workspace.

See Troubleshooting for startup, installation, and sync issues.