Meta Skill System

Problem

Meta Skill System addresses fragmented AI-agent workflows where teams spend time deciding whether a domain should exist, how to decompose a process, how to encode domain knowledge, and how to execute tasks consistently. It is not a vertical plugin; it is a self-referential meta-level execution framework that turns vague requirements into inspectable, reusable delivery paths and provides actionable decision rationale.

How It Works

The skill organizes M0 through M9 into ten domains. It starts with domain assessment: boundary, justification, eliminate/refactor/retain. Then workflow restructuring identifies core, calibration, coordination, and validation steps, compressing them into an IPO primitive chain with no more than five primitives. Next, payload generation produces catalog, requirements, and exemplars. Finally, a three-axis execution model handles tasks: the execution axis plans S/C/A/O/I/G operations and pipelines, the content axis consumes checklists or samples, and the innovation axis applies direct use, improvement, transfer, or construction.

Boundary: use it for method building, process compression, and skill-package generation. It does not replace factual verification, business authorization, or domain-specific compliance checks. It requires verifiable citations and marks uncertain cases as to verify; sample-based output should borrow structure and style, not copy content.