How does it handle ambiguous requests?

If the system's confidence in a specific route is below 90%, it will proactively ask the user for clarification and present 3-5 likely options to ensure the correct skill is invoked.

How accurate is the routing system?

The Skill Router is designed to achieve over 90% accuracy by utilizing intent extraction, domain mapping, and continuous learning from user feedback stored in local cache.

What does the Skill Router do?

It acts as a central traffic controller that analyzes user requests and sends them to the most relevant specialized skills within the SpecWeave ecosystem, such as the architect-agent or increment-planner.

Can it handle complex tasks that require multiple skills?

Yes, for complex requests like 'Build a SaaS', the Skill Router recognizes the high complexity and routes the task to a role-orchestrator that coordinates multiple agents simultaneously.

SpecWeave Skill Router

Name: SpecWeave Skill Router
Author: anton-abyzov

byanton-abyzov

生産性とワークフロー

Orchestrates complex development workflows by intelligently routing ambiguous user requests to the most appropriate specialized skills.

概要

The SpecWeave Skill Router acts as an intelligent traffic controller for AI-driven development environments. By analyzing intent, domain, and complexity, it ensures that every user request—from building new features to optimizing database queries—is handled by the correct specialized agent. It utilizes a sophisticated four-phase routing algorithm—Intent Classification, Skill Mapping, Confidence Assessment, and Clarification—to maintain over 90% routing accuracy. This skill is essential for managing multi-skill orchestration and resolving ambiguity in natural language prompts, ensuring developers get the right assistance without manual skill selection.

主な機能

High-accuracy intent classification using primary action and domain analysis
Interactive clarification system for requests with less than 90% confidence
Multi-skill orchestration for complex, cross-functional project requirements
Automatic performance tracking and machine learning feedback loops
Context-aware priority matrix for brownfield vs. greenfield development
0 GitHub stars

ユースケース

Resolving ambiguous prompts like 'help with auth' into specific planning or implementation tasks
Orchestrating full-stack project builds by coordinating architecture, backend, and frontend agents
Automatically selecting the correct documentation or testing skill based on file patterns

概要

主な機能

High-accuracy intent classification using primary action and domain analysis
Interactive clarification system for requests with less than 90% confidence
Multi-skill orchestration for complex, cross-functional project requirements
Automatic performance tracking and machine learning feedback loops
Context-aware priority matrix for brownfield vs. greenfield development
0 GitHub stars

ユースケース

Resolving ambiguous prompts like 'help with auth' into specific planning or implementation tasks
Orchestrating full-stack project builds by coordinating architecture, backend, and frontend agents
Automatically selecting the correct documentation or testing skill based on file patterns