autonomous-loop

Autonomous iterative development is orchestrated through planning, building, and status checking cycles. This Skill manages project specifications, implementation plans, and code generation through deterministic context loading and subagent delegation.

Autonomous multi-stage refactoring uses gap analysis and task prioritization to create implementation plans. It executes tasks one by one in a loop, ensuring focus and clear progress while handling production-quality code implementation.

Deterministic context loading fixes the project specifications and existing code state to ensure consistent autonomous code generation. It delegates execution to subagents, maintaining state persistence and integrated testing throughout the development loop.

Subagent delegation manages complex software engineering tasks by breaking them down into prioritized implementation plans. The system uses dual-condition exit gates to verify completion before moving to the next development cycle.

An autonomous loop is needed when a project requires significant refactoring and feature additions without constant human oversight. It manages the entire development process from initial planning to final testing, reporting progress at each step.

Start the autonomous development loop by invoking the /ralph command. This initiates the iterative cycles of planning, building, and status checking, reducing the need for constant human oversight during production-quality code implementation.