loop-driven-engineering

Managing complex multi-file refactors requires coordinating nested loops across code, deliverables, methods, and reasoning. This skill dispatches specialist sub-skills at the right moments, enforcing a fixed five-iteration budget with escalation rules to handle non-convergence and ensure disciplined progress.

Coordinating incident response across multiple code files is handled by orchestrating four nested loops that manage code, deliverables, methods, and reasoning chains. It enforces governance rules and ensures a current doc-level mental model for traceable closure.

Handling non-convergence when iterating on risky engineering tasks is managed by a hard five-iteration budget (K_MAX = 5). When iterations exhaust this budget, the system applies specific escalation rules to prevent endless loops and ensure task completion.

Nested loop coordination for cross-file bugfixes is needed for non-trivial tasks requiring synchronized progress across code, deliverables, methods, and reasoning. It maintains a synced mental model and dispatches sub-skills to manage complexity that simple linear execution cannot handle.

Loop-driven engineering works for tasks requiring strict audit trails by enforcing governance rules and generating traceable closure artifacts. These artifacts document the nested loop iterations across code and reasoning, providing clear handoff materials for audit purposes.