AI Production Relations Refactor

Problem it addresses

Multi-level reporting, coordination meetings, and formal approvals are often not required by the work itself. They are compensation relationships created by slow information flow, fragmented systems, and institutional constraints. This skill helps identify which roles, workflows, and approvals can be removed, which correction loops must remain, and how to compress a complex organization into an AI-assisted one-person execution chain.

How it works

• Decompose: map hierarchy, roles, information nodes, and coordination relationships, then ask why each relationship exists.
• Eliminate: remove transmission relationships, coordination relationships, and format relationships, while reducing verification relationships to critical checkpoints.
• Reorganize: combine retained core relationships and calibration relationships into I/P/O collaboration primitives, usually with primitive count ≤ 5.

It separates post-hoc verification from in-process calibration. Compliance, safety, and quality checks cannot be skipped, while stage reviews and feedback loops should remain as correction points inside a primitive so problems are not pushed to final delivery. The skill also provides an R0-01 → R0-06 execution chain and validation checks such as production completeness, efficiency preservation, and compliance retention.

Boundaries

It fits teams, projects, platforms, DAOs, and other multi-person collaboration structures, and is useful for shortening complex workflows. Legal, safety, and quality-related verification relationships must remain; fully automatic AI steps need fallback review; user-provided business constraints and calibration points override template defaults.