Audit Skill Lifecycle

Deep semantic validation of how skills interconnect, what they load, what they produce, and whether the resulting call graph forms a coherent, non-contradictory workflow.

Purpose

This audit traces every outbound reference from skills, agents, commands, and data files within a plugin to evaluate whether the resulting graph is coherent, complete, and non-redundant. It performs semantic validation beyond structural checks, answering whether the wiring makes sense given what each component claims to do.

This is NOT structural validation (that's what NamespaceReferenceValidator and /plugin-creator:assessor do). This is semantic validation of the skill lifecycle itself.

When to Use

Invoke this skill when:

• Preparing plugin for marketplace submission (semantic quality gate)
• Debugging why skill workflows fail to achieve stated goals
• Identifying instruction contradictions across related skills
• Finding duplicated content that should be extracted to shared references
• Detecting circular skill loading that could cause infinite recursion