What problem does it solve? In an ecosystem with dozens of AI skills, choosing which skill to activate for a given request is error-prone: ambiguous prompts trigger the wrong skill, or too many skills activate at once and waste context tokens. This Skill acts as the entry point for every substantive request, classifying intent and selecting the minimum sufficient skill set. ## Core Features & Use Cases - Intent classification and routing: Classifies requests as simple, complex, or ambiguous, then maps them to skills via semantic meaning rather than literal keyword matching, with confidence scoring thresholds. - Two-level domain routing: Routes first to a domain node (L1), then to a minimal skill set inside that node (L2), loading only one node manifest instead of every SKILL.md. - Multi-skill coordination: Defines handoff protocols, parallel activation rules, and typed-context passing between skills, with validation of combined outputs. - Use Case: A user asks to "build a validated n8n workflow and remember the session" — the router activates n8n-orchestrator and continuation-memory in parallel, then validation-mesh on the combined output, instead of loading all 28 skills. ## Quick Start Ask the assistant to route an ambiguous or multi-part request, for example: "Route this request: build a validated automation workflow and save our session state."
