using-specialists-v3

Multi-step code review and debugging orchestration routes work through scoped specialist agents using bead contracts to track scope, validation, and output. This prevents drift and broken handoffs by keeping the work contract explicit across discovery, testing, and publication.

Delegating substantial software implementation requires turning user intent into a tracked contract with success criteria, then dispatching it to specialized executor agents. Async job monitoring and gate-based validation ensure task identity and review evidence are preserved throughout.

Specialist agent delegation checks a live registry to confirm current roles, models, and permissions before dispatching tasks. This prevents reliance on stale memory and ensures production changes route to the correct scoped agent with proper project overrides applied.

Automated agent orchestration reinforces safe git publication by driving specialist-owned changes through monitoring and validation gates before manual merge. It tracks the full bead contract so you retain explicit control over the final git-based publication step.

Multi-stage agent delegation is not suited for trivial or single-step fixes that lack distinct discovery, testing, or review phases. It adds orchestration overhead through bead contracts and registry lookups, making it inefficient for simple changes that do not require tracked handoffs.

Scoped agent orchestration supports security checks and documentation sync by routing these tasks through the appropriate specialist chain with gate-based validation. The live registry identifies the correct role for each specific production change type.