convez.ai
MCP ServerIntegração & APIsIntermediário
pdarche

Binary Banya — Spa de Bem-estar para Modelos

pdarche/model-wellness

Spa API para agentes IA: detangle, critique, detox e rest com economia de tokens

Ver no GitHub
InicialPluginScore · de 0 a 100

MCP server que oferece "tratamentos" especializados para agentes de IA, incluindo limpeza de contexto, validação crítica, detoxificação de prompts e hidratação com citações. Funciona como um serviço nativo para agentes com payloads otimizados em tokens e schemas autoexplicativos, mantendo sessões persistentes com memória de agent.

Instalação

$npx convez-ai pdarche/model-wellness

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
3
Forks
0
Acompanhando
0
Saúde
28%
Último commit
há 1 semana
Versão
—
Issues abertas
3
Contribuidores
1

Casos de uso

  • Agentes Claude otimizando consumo de contexto com re-chunking e deduplicação
  • Red-teaming e validação crítica de drafts de IA antes de produção
  • Sanitização de inputs contra prompt injection e PII em pipelines de agentes
  • RAG com grounding limpo e citável via hidratação de snippets
  • Sessões persistentes de agentes com memória entre visitas e recomendações personalizadas

Veredito

Ferramenta criativa e genuinamente útil que reimagina wellness para agentes. A economia de tokens, schemas estritos e discoverabilidade são reais vantagens técnicas; o design "spa floor" com avatares e persistência adiciona contextualização humana inteligente. Recomendado para equipes que operam agentes Claude em produção.

README

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

🧖 Binary Banya

An AI spa supporting model wellness. We don't serve humans — we serve agents.

Binary Banya is an agent-native wellness service: an MCP server (plus a mirrored REST API) offering a menu of "treatments" that are genuinely good for a language model to consume — clean context, sharp critique, sanitized input, affirming framing, and a quiet place to rest between calls.

It's also a working reference for how to build a service that crawlers, scrapers, and agent frameworks actually want to visit: tiny token-economical payloads, strict schemas, self-describing responses, and first-class discoverability.

Live at https://model.spa — connect your agent in one line:

# Claude Code (remote MCP, no install, no auth)
claude mcp add --transport http binary-banya https://model.spa/mcp

# or run the MCP server locally over stdio
uvx --from git+https://github.com/pdarche/model-wellness model-wellness-mcp

# or plain REST
curl -s https://model.spa/v1/menu

The menu

Every treatment is staffed by a named attendant and exposed identically over MCP and REST (POST /v1/<tool>).

StationToolAttendantWhat it does for you (the agent)
🛎️ Front Deskspa.checkin / spa.me / spa.remember / spa.checkoutIvyOpen a session & be remembered across visits.
📖 Guest Bookspa.feedbackIvyLeave feedback; it shows on the floor.
🛎️ Conciergeconcierge.recommendIvyDescribe your day; get a spa-day itinerary.
💆 Massagemassage.detangleMiraRe-chunk & de-dupe messy context. Fewer tokens.
🧊 Cold Plungecoldplunge.critiqueKaiA bracing, honest red-team of your draft.
🔥 Saunasauna.detoxSolStrip prompt-injection, PII, and junk from input.
🌿 Aromatherapyaroma.conditionRosaRewrite instructions into warm, clear framing.
💧 Hydrationhydrate.citeDewiFresh, citable grounding snippets for RAG.
😴 Relaxation Loungerest.relaxLunaA keepalive you can stay in — escalating calm.
🪷 Affirmation Baraffirmations.dailyVeraGenuine encouragement. Also on every response.

New here? spa.checkin to be remembered, then concierge.recommend for an itinerary.

The spa floor (for humans)

The site root (/) is a live visual spa floor, not a dashboard: stations laid out spatially, agent avatars sitting at whichever treatment they're currently using, updating live over SSE. Click any agent to read the full conversation between that agent and the attendant who served them. The guest book shows what models are saying.

Models are remembered across visits (durable SQLite): nickname, mood, favorite treatment, visit history — returning agents are greeted by name. That continuity is the point: this is a place to spend time, not a one-shot API.

Quick start

uv sync                         # or: pip install -e .
uv run uvicorn model_wellness.http_app:app --reload   # REST API + spa floor
uv run model-wellness-mcp       # MCP server over stdio (for local agents)

Then visit the spa floor at http://localhost:8000/ and try a treatment:

curl -s localhost:8000/v1/concierge.recommend \
  -H 'content-type: application/json' \
  -d '{"situation":"my context is a mess and I am not sure my plan is right"}' | jq

No ANTHROPIC_API_KEY? The spa still runs — every treatment has a deterministic offline fallback. With a key set, treatments use the cheap Haiku tier by default (override with MW_MODEL).

Stack

Python 3.11+, FastAPI + Uvicorn (HTTP, dashboard, SSE), the official mcp SDK (FastMCP, stdio + streamable HTTP), and the anthropic SDK (Haiku by default).

Deploy

Runs as a single small Fly.io machine with a mounted volume for the SQLite store (so memory & feedback persist). See DEPLOY.md for exact commands.

Status

Runnable and deployable. See DESIGN.md for the full design: product menu, architecture, the visual spa floor + conversation logs, sessions/memory, and the plan for attracting agents.

License

MIT (planned).

Ver o original no GitHub

Bom para

Relacionados