agent-selection

Agent selection uses decision-tree logic to map task characteristics—framework dependency, domain, project type, and quality concerns—to the optimal specialized agent. Analyze whether the task involves React, TypeScript, Vite, security audits, performance profiling, code review, or architecture refactoring, then delegate with mandatory context references to docs.

Agent selection handles framework-specific work (React, TypeScript, Vite, Vitest), domain-specific tasks (security audits, performance profiling, API documentation), project types (animation design, UI polish, plugins), and quality concerns (code review, testing strategy, architecture refactoring) through specialized agent routing.

After validating expertise alignment and scope, delegation uses the Task tool with context-rich instructions and mandatory references to relevant documentation (docs/architecture.md, docs/api.md, docs/testing.md) to ensure the selected agent has all necessary context.

Use agent selection when tasks span multiple domains, require specialized expertise verification, or involve complex orchestration across framework-specific, quality, and architectural concerns. It reduces misalignment by systematically matching task characteristics to expert capability before delegation.

Yes, agent selection implements multi-domain orchestration logic to map complex tasks that span frameworks, quality concerns, and specialized domains to the correct agent or agent combination through decision-tree validation.