devflow

End-to-end software development workflows are automated by orchestrating exploration, plan refinement, spike prototyping, implementation, and verification with built-in quality gates. Auto mode enables autonomous task execution while maintaining logging and isolation gates for reliable outcomes.

Spike prototyping in structured development workflows is a coordinated phase where preliminary implementation validates technical assumptions before full execution. It sits between plan refinement and implementation, ensuring that downstream WORKLOG and DR loops operate on verified foundations.

A clear task specification is required for autonomous development workflow orchestration. This Skill applies when a user has a defined task and wants autonomous execution with quality gates, managing artifacts like PLAN.md, WORKLOG.md, exploration.md, and Decision Records.

Decision record handling works by capturing architectural and implementation choices during WORKLOG loops to ensure full traceability. The orchestration manages these DR artifacts alongside implementation progress, preventing misaligned plans and fragmented task tracking.

The best way to maintain traceability across multi-stage reviews is through structured artifact management using PLAN.md, WORKLOG.md, and Decision Records. Coordinated end-to-end orchestration logs each stage from discovery to verification, preventing inconsistent quality gates.

You should not use auto mode for development workflow orchestration when a task specification is unclear or when manual oversight is required. The orchestration depends on clear specs to trigger autonomous execution with quality gates, isolation gates, and DR handling effectively.