kiro-impl

Autonomous TDD task implementation is automated by dispatching sub-agents to execute pending tasks defined in tasks.md. The workflow validates completion criteria using parent-provided tests, builds, and smoke commands to ensure auditable progress tracking.

A KIRO-like workflow is a two-mode process that coordinates multi-agent AI flow to reduce manual orchestration. It operates in either autonomous sub-agent dispatch or manual execution mode, enforcing strict auditable processes with status tracking documented in tasks.md.

Yes, TDD workflow automation requires projects to have tasks defined in tasks.md with linked spec sections in requirements.md and design.md. These files provide the approved task definitions and design specifications needed for sub-agent dispatch and validation.

Yes, you can manually select specific tasks for execution by providing them in the main context. The workflow supports a manual task selection mode that runs specific user-provided tasks instead of automatically processing the next pending task.

Autonomous task execution tracks progress and ensures auditability by documenting decisions and status in tasks.md with explicit commit messages and boundaries. This enforces strict process validation against parent-provided tests and smoke commands.

Sub-agent dispatch for TDD task implementation is limited to projects with tasks already defined in tasks.md and linked specifications in requirements.md and design.md. It requires parent-provided tests, builds, and smoke commands to validate completion criteria.