taskyou

TaskYou is a sophisticated orchestration tool designed for developers who need to manage complex, multi-stage projects using AI agents. By integrating directly into the terminal via a robust CLI and a clean Kanban UI, it bridges the gap between high-level task planning and hands-on code execution. Each task is isolated within its own git worktree, ensuring that concurrent processes—such as background development, testing, and debugging—do not conflict with one another. The system is built for power users who leverage AI for code generation, architectural exploration, and bug fixing.

• Kanban board interface for visual tracking of tasks across backlog, processing, blocked, and done states.

• Isolated environment execution using git worktrees to prevent dependency and file conflicts.

• Pluggable AI backend architecture allowing users to swap between Claude Code, OpenAI Codex, Gemini, Pi, and open-source models.

• Intelligent codebase context caching that stores project structure, tech stack, and coding patterns to improve token efficiency and agent accuracy.

• Real-time interaction capabilities, including the ability to send direct input to shell processes, confirmation of prompts, and monitoring of live execution logs.

• Event-driven hooks that trigger scripts on status changes, enabling integration with desktop notifications, email reports, or CI/CD pipelines.

• Comprehensive CLI for full automation, supporting JSON output for seamless integration with tools like jq.

• Users should start by surveying the current board state using the board subcommand to identify high-priority or blocked tasks.

• To optimize performance, agents utilize taskyou_get_project_context to retrieve previously cached codebase insights.

• When a task enters a blocked state due to a prompt or missing information, users can provide feedback or direct key-press inputs via the input command.

• The system supports SSH remote access, making it suitable for managing tasks across multiple machines or in cloud environments.

• Deployment requires Go 1.24.4+ and assumes familiarity with standard git workflows and terminal-based TUI interfaces.