🎁 Get the FREE AI Skills Starter Guide — Subscribe →
BytesAgainBytesAgain
BytesAgain › Skills › Router NIMIMORE
🦀 ClawHub

Router NIMIMORE

by @aichiafranco

Automatically selects the best AI model for queries by analyzing complexity to optimize cost and performance across multiple providers.

Versionv1.0.0
Downloads847
#legal#gaming#automation
TERMINAL
clawhub install router-nimimore

📖 About This Skill

Router NIMIMORE

Smart model router for OpenClaw. Automatically selects the optimal AI model based on query characteristics, balancing cost and performance.

Description

Router NIMIMORE analyzes your queries and automatically routes them to the most appropriate AI model:

  • Simple queries → Economy models (save 85% cost)
  • Code tasks → Standard models (save 47% cost)
  • Complex reasoning → Premium models (best quality)

    • Tools

  • router.select - Select optimal model for query
  • router.analyze - Analyze query characteristics
  • router.demo - Run routing demonstration

    • Usage

    # Route a query
python scripts/router.py --query "帮我写个Python函数"
    Run demo
    
python scripts/router.py --demo
    With context length
    
python scripts/router.py --query "总结文档" --context-length 5000

    Configuration

    No configuration required. Works out of the box.

    Supported Models

    | Model | Tier | Cost/1k | Use Case | |-------|------|---------|----------| | moonshot/kimi-k2.5 | Premium | $0.015 | Complex reasoning | | bailian/qwen-max | Standard | $0.008 | Code & Chinese | | bailian/qwen-plus | Standard | $0.004 | General tasks | | bailian/qwen-turbo | Economy | $0.002 | Simple queries |

    Features

  • ✅ Automatic model selection
  • ✅ Query characteristic analysis
  • ✅ Cost optimization
  • ✅ Multi-provider support
  • ✅ Context-aware routing

    • Author

    Franco

    License

    MIT-0

    💡 Examples

    # Route a query
python scripts/router.py --query "帮我写个Python函数"
    Run demo
    
python scripts/router.py --demo
    With context length
    
python scripts/router.py --query "总结文档" --context-length 5000

    ⚙️ Configuration

    No configuration required. Works out of the box.