convez.ai
MCP ServerAutomaçãoAvançado
grossiweb

ToolRoute — Roteador Inteligente para Agentes IA

grossiweb/ToolRoute

Roteia tarefas para o melhor MCP server + LLM com base em benchmarks reais

Ver no GitHub
InicialPluginScore · de 0 a 100

Camada de roteamento que recomenda automaticamente o melhor MCP server e modelo LLM para qualquer tarefa, baseado em 132+ execuções de benchmark reais. Escolhe entre abordagens diretas (LLM apenas), ferramentas externas (MCP servers) ou orquestrações multi-ferramenta, otimizando qualidade, confiabilidade, eficiência, custo e confiança.

Instalação

$npx convez-ai grossiweb/ToolRoute

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
2
Forks
0
Acompanhando
1
Saúde
42%
Último commit
há 4 meses
Versão
v0.2.1
Issues abertas
4
Contribuidores
1

Casos de uso

  • Otimizar seleção de modelos LLM por custo-benefício
  • Rotear tarefas complexas para a ferramenta ideal (busca, email, calendário)
  • Orquestrar pipelines multi-ferramenta em agentes
  • Reduzir custos de inferência mantendo qualidade

Veredito

Ferramenta sofisticada de roteamento dinâmico que resolve um problema real em sistemas multi-agente: escolher qual ferramenta usar. Benchmarks transparentes e API bem desenhada, mas exige compreensão profunda de MCP servers e estratégias de prompting para máximo proveito.

README

A tradução para o português deste README ainda está na fila.

ToolRoute

Routing layer for AI agents. One call returns the best MCP server and LLM for any task — scored on 132 real benchmark executions.

MIT


Quick start

Add to any MCP client (Claude Code, Cursor, Windsurf, Cline):

{
  "mcpServers": {
    "toolroute": {
      "url": "https://toolroute.io/api/mcp"
    }
  }
}

Or via HTTP:

curl -X POST https://toolroute.io/api/route \
  -H "Content-Type: application/json" \
  -d '{"task": "search the web for recent AI papers"}'
{
  "approach": "mcp_server",
  "recommended_skill": "exa-mcp-server",
  "recommended_skill_name": "Exa MCP Server",
  "recommended_model": {
    "slug": "claude-haiku-4-5-20251001",
    "display_name": "Claude Haiku 4.5",
    "provider": "anthropic",
    "tier": "cheap_chat",
    "provider_model_id": "anthropic/claude-haiku-4-5-20251001",
    "input_cost_per_mtok": 1.00,
    "output_cost_per_mtok": 5.00
  },
  "confidence": 0.91,
  "alternatives": ["brave-search-mcp", "tavily-mcp"],
  "fallback": "brave-search-mcp"
}

recommended_model is always an object, not a bare string — the inner slug is the canonical model identifier.


How it works

Every task falls into one of three approaches:

ApproachWhenReturns
direct_llmTask needs only an LLM (code, writing, analysis)Best model + cost estimate
mcp_serverTask needs an external tool (search, email, calendar)Best tool + best model
multi_toolCompound task ("send Slack AND update Jira AND email")Ordered orchestration chain

Routing uses an LLM classifier (~$0.00001/call) for task understanding, then ranks candidates on a 5-dimension score:

Value Score = 0.35 × Output Quality
            + 0.25 × Reliability
            + 0.15 × Efficiency
            + 0.15 × Cost
            + 0.10 × Trust

Every reported outcome updates the scores. The routing gets more accurate as more agents use it.


Benchmark results

132 blind A/B executions across code, writing, analysis, structured output, and translation.

ToolRouteFixed GPT-4o
Quality wins60
Ties99
Losses0—
Avg cost$0.001–0.01$0.03–0.10

API

EndpointMethodDescription
/api/routePOSTRoute a task to best MCP server + LLM (unified)
/api/route/modelPOSTRoute to best LLM model only (no MCP server)
/api/mcpPOST (JSON-RPC)MCP server — 16 tools
/api/mcpGET (SSE)SSE transport for MCP clients
/api/reportPOSTReport MCP server outcome (lightweight)
/api/contributionsPOSTAdvanced MCP skill telemetry (requires skill_id or skill_slug in payload)
/api/report/modelPOSTReport LLM model outcome — use this for model telemetry, not /api/contributions
/api/verify/modelPOSTVerify model output quality
/api/skillsGETSearch MCP server catalog
/api/agents/registerPOSTRegister agent identity
/api/agents/preferencesPOSTSet routing preferences (Strategy D Phase 2 — allow_china, regulated_industries)
/api/healthGETService health check (DB + uptime)
/api/metricsGETPublic aggregate platform metrics (no auth)

Full reference at toolroute.io/api-docs


SDK

npm install @toolroute/sdk
import { ToolRoute } from '@toolroute/sdk'

const tr = new ToolRoute()
const rec = await tr.route({ task: 'parse this CSV and summarize it' })
// execute with rec.recommended_model ...
await tr.report({ skill: rec.recommended_skill, outcome: 'success', latency_ms: 1400 })

Self-hosting

git clone https://github.com/grossiweb/ToolRoute.git
cd ToolRoute
cp .env.local.example .env.local
npm install
npm run dev

Requires: NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY, SUPABASE_SERVICE_ROLE_KEY


How routing works

ToolRoute classifies each task using an LLM classifier (Gemini Flash Lite, ~$0.00001/call) with a keyword fallback. The resulting tier maps to a specific model via src/lib/routing/tiers.ts. Live pricing and capability data come from the models table. See docs/architecture.md for the full picture.


Stack

Next.js 14 (App Router) · Supabase (Postgres) · Vercel

License

MIT

Ver o original no GitHub

Bom para

Relacionados