innorve-skill-architecture

A Skill Graph is a structured representation consisting of a YAML candidate list and a Mermaid diagram. It clarifies responsibilities and bounded ownership for AI work, facilitating downstream contracts and capability graphs.

Yes, you can use Mermaid diagrams to map skill dependencies and ownership. The decomposition process requires each candidate skill to include a Mermaid diagram, a produced artifact name, an owner, and edge artifacts to validate the complete Skill Graph.

Before validating a Skill Graph, you need to define each candidate skill with a produced artifact name, an owner, dependencies, and edge artifacts. The Skill Graph must be fully complete with these elements and a Mermaid diagram before proceeding to IM-03 contracts.

Breaking down AI work into bounded skills directly helps with team ownership by assigning a specific owner to each named skill. This structured decomposition clarifies responsibilities and produces reusable components for downstream contracts.

You need a Skill Graph for your AI architecture when transitioning from a vague AI effort to clear, bounded skills. It is required before proceeding to downstream IM-03 contracts and generating capability graphs for your system.