Run Getting Things Done With AI-Powered Planning

Implements the complete Getting Things Done methodology as an integrated workflow. It captures tasks, routes them intelligently, organizes them by context and project, plans your day based on energy levels, and conducts reflective weekly reviews. This removes the friction of jumping between multiple tools and the cognitive load of manually processing captures into the GTD system.

Once set up, you work through a five-phase daily and weekly cycle: 1. Capture tasks using \capture [item], and the AI auto-processes obvious single actions directly to projects while sending unclear items to inbox. 2. Process inbox items during your workflow by clarifying scope and context. 3. Organize tasks by active projects and contexts (@work-code, @home-calls, etc.). 4. Plan your day with \daily, which pulls high-priority tasks, maps them to energy levels, and creates time blocks. 5. Conduct weekly reviews with \weekly to reflect on accomplishments and plan the next week. Output: a single source of truth for all tasks, organized by project and context, with daily and weekly reflection logs.

Captures, processes, and plans all in one system, so task context and project state flow through each phase without information loss.

Requires Claude Code as the AI runtime for command processing. Just (https://github.com/casey/just) is used as the task runner to orchestrate commands. Tod (https://github.com/alanvardy/tod) enables CLI-based Todoist sync. icalBuddy (macOS only via Homebrew) integrates calendar events. The system uses markdown files for storage and relies on AI-driven processing for smart task routing.

Requires Claude Code, Just, Tod (Todoist CLI), and icalBuddy for full functionality. The system depends on consistent daily and weekly review discipline to maintain effectiveness. macOS only for calendar integration via icalBuddy.