The Meta-Theory skill (Meta Arsenal Dispatcher) provides governance and orchestration for multi-agent development and review workflows. It structures work into an 8-stage spine (Critical → Fetch → Thinking → Execution → Review → Meta-Review → Verification → Evolution), enforces clarity checks, and routes execution to capability-matched owners rather than executing risky tasks itself. It's aimed at governance, architecture decisions, multi-file development coordination, and delivery verification.
Use this skill when you need formal agent governance: clarifying intent and success criteria, discovering capabilities, planning multi-agent or multi-file work, performing security/quality reviews, or verifying releases and fixes. Also use it for architecture decisions, agent design, and debugging runs that require structured review and verification.
Compatible with governance and agent orchestration toolchains (meta-conductor, meta-prism, Codex/Claude adapters, OpenClaw dispatch patterns). The skill is runtime-agnostic but expects access to shell, filesystem, browser, and memory tools as listed in its frontmatter.
This skill has not been reviewed by our automated audit pipeline yet.
