agent-gtd
Executive function system for AI agents: GTD-style task management, error/learning capture, and periodic self-review using Taskwarrior. Use when: (1) tracking multi-step goals or internal agent tasks, (2) logging command failures, user corrections, or missing capabilities, (3) running a self-improvement or review loop, (4) managing continuity across sessions, (5) triaging a backlog of agent work. Reference implementation uses Taskwarrior; patterns are portable to other task backends.
