SmartAgent 🤖
Biblioteca Python leve para criar agentes de IA com execução determinística em 3 fases.
Características ✨
- Multi-Provider: Suporta Groq, OpenAI, Gemini, Grok, Ollama e Llama
- Sem Schemas: Não precisa de Pydantic ou definições complexas
- 3 Fases: Análise → Execução → Resposta
- Econômico: Minimiza uso de tokens
- Simples: Menos de 500 linhas de código
Instalação
# Instalar dependências
pip install requests
Uso Rápido
from agent import Agent
# Criar agente
agent = Agent(provider="groq")
# Registrar ferramentas
@agent.tool
def get_products(max_price=100):
return [{"nome": "Mouse", "preço": 50}]
# Executar
response = agent.chat("Quais produtos baratos?")
print(response)
Instruções Customizadas
Você pode adicionar instruções personalizadas ao agente:
agent = Agent(
provider="groq",
info="""
Você é um assistente especializado em e-commerce.
- Sempre sugira produtos relacionados
- Use tom amigável e profissional
- Destaque promoções quando disponíveis
"""
)
Providers Suportados
- Groq:
Agent(provider="groq", api_key="...") - OpenAI:
Agent(provider="openai", api_key="...") - Gemini:
Agent(provider="gemini", api_key="...") - Grok:
Agent(provider="grok", api_key="...") - Ollama:
Agent(provider="ollama")(local) - Llama:
Agent(provider="llama", api_key="...")
Variáveis de Ambiente
# Provider default global (opcional)
export SMARTAGENT_PROVIDER="groq"
export GROQ_API_KEY="your-key"
export OPENAI_API_KEY="your-key"
export GEMINI_API_KEY="your-key"
export XAI_API_KEY="your-key"
export LLAMA_API_KEY="your-key"
# Modelo global (novo)
export SMARTAGENT_MODEL="modelo-ai"
# Modelo por provider (opcional)
export SMARTAGENT_OPENAI_MODEL="gpt-4o-mini"
export SMARTAGENT_GROQ_MODEL="qwen/qwen3-32b"
# Rede (opcional)
export SMARTAGENT_TIMEOUT="30"
export SMARTAGENT_RETRIES="2"
# Compatibilidade legada (ainda suportado)
export LLM="modelo-ai"
Compatibilidade (1-2 versões)
model=\"groq\"(estilo antigo) ainda funciona, mas o recomendado éprovider=\"groq\".- As variáveis
*_API_KEYcontinuam suportadas. LLMcontinua suportada como fallback para modelo global.
Arquitetura
- Analyzer: Determina quais ferramentas executar
- Executor: Executa ferramentas com parâmetros
- Responder: Gera resposta humanizada
Exemplo Completo
Veja examples/minimal_agent.py
Metadata
Release files for smartagent-sf 0.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| smartagent_sf-0.2.0.tar.gz | 16.4 kB | Without provenance | Details for smartagent_sf-0.2.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| smartagent_sf-0.2.0-py3-none-any.whl | Python 3 | none | any | Details for smartagent_sf-0.2.0-py3-none-any.whl |
Total release size: 36.6 kB
Release files / smartagent_sf-0.2.0.tar.gz
| Download URL | smartagent_sf-0.2.0.tar.gz |
|---|---|
| Size | 16.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3f6a73ee5fa40cdc633c0ff224686aa094d5eb86bd4011cc41263fcdfc8109cf
|
|
BLAKE2b-256 checksum How to use checksums |
d63024b0990e70ccd80adebdd19f150382472bce6393c85c02ebef0d629ea2f1
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Release files / smartagent_sf-0.2.0-py3-none-any.whl
| Download URL | smartagent_sf-0.2.0-py3-none-any.whl |
|---|---|
| Size | 20.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
63415fd355e2ca08fa38952743ba40aa168abddc17e28b9e3c88f3f12a59bcee
|
|
BLAKE2b-256 checksum How to use checksums |
322cca32d3e3a76c58b39495f754a25b707f54ae1c63e9de04813d1e08312437
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Security for smartagent-sf 0.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.