things-mac

Things-mac is a specialized integration for the OpenClaw personal AI assistant that provides full-featured command-line control over your Things 3 database. Designed exclusively for macOS, it bridges the gap between your AI assistant and your primary task management system. By interacting directly with the local Things 3 data store and utilizing the Things URL scheme, this skill allows users to maintain a structured task workflow entirely through natural language or automated triggers. Whether you are looking to query your current inbox, prioritize today's tasks, or inject new items into specific projects or areas, this tool serves as a reliable interface for seamless task synchronization.

• Perform comprehensive read operations including inbox retrieval, today's agenda, upcoming deadlines, and search across all tasks, projects, areas, and tags.

• Execute write operations such as adding new todos with metadata like notes, deadlines, scheduling dates, and project assignments.

• Update existing items by modifying titles, notes, tags, or moving tasks between lists and headings.

• Support for advanced task structure including checklist item creation and multi-line note handling via standard input.

• Integration with Things 3 URL scheme enables bringing the application to the foreground and performing dry-run operations for safe verification of commands.

• High flexibility in configuration, allowing custom database paths and auth tokens for secure, authenticated updates.

• Ensure the calling application (such as Terminal or OpenClaw) has Full Disk Access granted in macOS System Settings to read the local Things database.

• While read operations are handled via local database access, write operations require the Things URL scheme and an optional auth token for secure modifications.

• Note that some destructive actions like 'delete' are not natively supported; users should rely on marking items as completed or canceled.

• The tool is strictly limited to macOS environments where Things 3 is installed.

• Use the --dry-run flag during initial setup or testing to preview commands without executing them against your live task database.