smart-model-routing
SolidDynamic model selection based on task complexity scoring. Replaces static model mappings with a weighted signal system that picks Opus, Sonnet, or Haiku-class speed per task. Works with agent-assignment-matrix.md.
Install
Quality Score: 86/100
Skill Content
Details
- Author
- vibeeval
- Repository
- vibeeval/vibecosystem
- Created
- 2 months ago
- Last Updated
- 1 months ago
- Language
- C#
- License
- MIT
Integrates with
Similar Skills
Semantically similar based on skill content — not just same category
model-routing
Vendor-neutral routing guide for choosing the right model tier by task type. Mechanical work uses a smaller/faster model; implementation uses a standard model; architecture, security, and release audit use the most capable model.
model
Model selection strategy and routing. Use when choosing between models for different task types, subagent configurations, or optimizing token cost vs quality tradeoffs.
smart-routing
Complexity-based task routing with Q-Learning optimization, Agent Booster WASM fast-path, and Mixture-of-Experts model selection.
cost-routing
Top-level dispatcher that classifies every incoming request into scout / coder / architect tiers BEFORE any tool call. Routes Read, Grep, Glob, file-search, symbol-lookup, "where is X", and "list files matching Y" to haiku-scout. Routes known-location Edit, Write, multi-file refactor, test authoring, and bounded code changes to sonnet-coder. Reserves the main opus context for ambiguous design questions, ADRs, and tradeoff analysis. Use whenever a request lands in the main context and might involve file IO, code search, code edits, or design reasoning — which is almost every turn.
cost-aware-pipeline
Cost-aware LLM pipeline patterns for optimal model routing, narrow retry strategies, and prompt caching. Reduces API costs 40-70% through intelligent model selection, targeted retries, and cache-friendly prompt structures. Use when: (1) Building multi-model pipelines, (2) Optimizing API costs, (3) Designing retry strategies for LLM calls, (4) Implementing prompt caching, (5) Choosing between haiku/sonnet/opus for sub-tasks.