ralphinho-rfc-pipeline

The ralphinho-rfc-pipeline skill provides a structured framework for managing complex software changes that are too extensive for a single-agent workflow. It leverages an RFC-driven multi-agent directed acyclic graph (DAG) execution pattern to decompose large features into independently verifiable work units. This approach ensures that each unit can be implemented, tested, and validated separately, reducing the risk of integration failures and improving overall development reliability. By orchestrating work across multiple agents and incorporating quality gates, merge queues, and recovery strategies, it addresses the challenges of scaling collaborative software development and maintaining code integrity.

Key features of the ralphinho-rfc-pipeline include a multi-stage pipeline starting from RFC intake and DAG decomposition to final system verification. Each work unit is defined with a clear specification including dependencies, scope, acceptance tests, risk level, and rollback plans. Units are categorized into complexity tiers based on the scope and risk of changes, ranging from isolated file edits to security-critical schema or performance modifications. The skill also provides a structured quality pipeline for each unit, covering research, implementation planning, coding, testing, review, and preparation for merge. Merge queue rules enforce dependency resolution, rebasing, and re-testing, while recovery procedures handle stalled units by evicting, snapshotting, narrowing scope, and retrying, ensuring continuous progress.

This skill is particularly useful for software engineers, project leads, and teams managing large-scale features that require rigorous validation and risk management. Ideal use cases include multi-file refactoring, critical system updates, or features that span multiple services or modules. By enabling decomposition into independently testable units and orchestrating integration systematically, ralphinho-rfc-pipeline facilitates efficient collaboration, reduces integration risks, and provides clear documentation of unit-level progress and outcomes, making it suitable for both development and quality assurance workflows.