What problem does it solve? Complex multi-domain tasks often get handled sequentially or delegated vaguely, producing slow execution and poor results. This Skill turns the AI into a coordinator that decomposes work, dispatches parallel worker agents, and synthesizes their findings into a unified output. ## Core Features & Use Cases - Structured Lifecycle: Follows a six-phase loop — decompose, classify, dispatch, monitor, synthesize, verify — with explicit concurrency rules for parallel reads and sequential writes. - Worker Prompt Discipline: Provides templates for research, implementation, and verification workers that forbid vague delegation like "based on your findings" and require file paths, line numbers, and scope boundaries. - Fork & Spawn Semantics: Defines when to fork context-inheriting agents versus spawning fresh specialists, plus anti-patterns such as fabricating worker results or skipping the synthesis phase. - Use Case: When asked to refactor an authentication module, the coordinator launches parallel research workers on the token logic and session handling, synthesizes findings, then dispatches sequential implementation workers and independent verification workers. ## Quick Start Ask the AI to coordinate a complex multi-file task using coordinator mode, for example: orchestrate a security audit and refactor of the authentication module with parallel research workers.
