pagent 0.1.1

Minimal OpenAI Chat Completions agent (session + tools).

pagent (English)

Language: 中文 | English

Lightweight async agent core for OpenAI-compatible Chat Completions.

Core Features

• Session: conversation message buffer (session += {...} / reset())
• LLM: thin stateless wrapper over AsyncOpenAI
• Providers: DeepSeek, Ollama, Vllm, Sglang
• @tool() / FunctionTool: convert Python callables into tool schemas
• Agent: multi-turn tool loop until no tool calls or max_turns reached
• Built-in tools: clock, region (locale/timezone hints, not GPS)

Install

pip install pagent

From source:

cd pagent
uv sync
pip install -e .

Quick Start

import asyncio
import os

from pagent import Agent, LLM, Session, tool


@tool()
def get_weather(city: str) -> str:
    """Return a fake weather summary for the city."""
    return f"It's sunny in {city} today."


async def main() -> None:
    if not os.getenv("OPENAI_API_KEY"):
        raise SystemExit("Please set OPENAI_API_KEY first.")

    agent = Agent(
        llm=LLM("gpt-4o-mini"),
        session=Session("You are a concise assistant. Use tools when needed."),
        tools=[get_weather],
        max_turns=8,
    )

    result = await agent.run("What's the weather in Xiamen?")
    print(result.content)
    print(agent.stats)


asyncio.run(main())

Environment Variables

• LLM(...) reads OPENAI_API_KEY by default
• DeepSeek(...) reads DEEPSEEK_API_KEY by default
• Local providers can use OLLAMA_API_KEY / VLLM_API_KEY / SGLANG_API_KEY (dummy key is used when missing)

DeepSeek

• Docs: DeepSeek API Docs
• base_url: https://api.deepseek.com
• default model: deepseek-v4-flash

from pagent import DeepSeek

llm = DeepSeek()  # or DeepSeek("deepseek-v4-pro")

Advanced fields can be passed through request_kwargs to chat.completions.create.

Local Deployment

If your server exposes OpenAI-compatible /v1/chat/completions, it should work directly:

• Ollama: http://127.0.0.1:11434/v1
• Vllm: http://127.0.0.1:8000/v1
• Sglang: http://127.0.0.1:30000/v1

from pagent import Ollama

llm = Ollama("llama3.2")

Maintainer: Publish to PyPI

Included workflow: .github/workflows/publish.yml (triggered by GitHub release publish).

Recommended: Trusted Publishing (OIDC).

• Docs: https://docs.pypi.org/trusted-publishers/

Notes

This library assumes OpenAI Chat Completions compatibility. If your backend API shape differs significantly, adapt at the gateway layer or override LLM.invoke.