What problem does it solve? Deciding when to split a research task across multiple AI agents is hard: delegating too eagerly wastes tokens and adds coordination overhead, while doing everything solo can miss coverage or blow the context budget. This Skill provides a disciplined gate, delegation design, and synthesis process for coordinator-worker research. ## Core Features & Use Cases - Delegation Gate: Applies explicit criteria to decide whether a task should be delegated or kept with a single strong agent, avoiding uneconomic fan-out. - Worker Brief & Output Schema: Provides templates for bounded worker assignments and compact evidence rows with status values like supported, conflicting, or not found. - Provenance-Aware Synthesis: Builds claim tables and evidence matrices before writing conclusions, preserving retrieval dates and unresolved uncertainty. - Use Case: When asked to verify 30 claims across dozens of sources, use this Skill to decide the split, assign 4-12 workers with source policies, then synthesize an auditable claim table with cost and latency comparison against a solo baseline. ## Quick Start Use the multiagent-research-coordinator skill to decide whether my research task should be delegated to workers and, if so, draft the worker briefs and synthesis plan.
