scope-oaf 0.2.0

OpenAgentFramework — minimal, fast, transparent AI agent framework

OAF — OpenAgentFramework

Minimal, fast, transparent AI agent framework for Python.

JSON-based tool calling, provider-agnostic (OpenAI + Anthropic), full prompt inspectability.

Install

pip install scope-oaf

Or from source:

pip install /path/to/oaf

Quick Start

import asyncio
from oaf import Agent, ToolRegistry

registry = ToolRegistry()

@registry.register
async def get_weather(city: str) -> str:
    """Get the current weather for a city."""
    return f"Weather in {city}: 72°F, sunny"

agent = Agent(model="gpt-4.1-nano", tools=registry)
response = asyncio.run(agent.message("What's the weather in Tokyo?", role="user"))
print(response.text)

Top-level API

from oaf import Agent, Hooks, Tool, tool, tool_method, BaseToolGroup, ToolRegistry

That's the full autocomplete list. Everything else lives in subpackages.

Subpackages

# LLM client, types, models, errors
from oaf.llmclient import LLMClient, Message, ChatResponse, Models, LLMError

# Context / memory
from oaf.context import BaseContext, ConversationalInMemory

# Prompt builders
from oaf.prompts import BasePromptBuilder, DefaultPromptBuilder

Tool Groups

from oaf import BaseToolGroup, tool_method, ToolRegistry

class MathTools(BaseToolGroup):
    """Mathematical operations."""
    name = "math"

    @tool_method
    async def add(self, a: int, b: int) -> str:
        """Add two numbers."""
        return str(a + b)

registry = ToolRegistry()
registry.register_group(MathTools())

Providers

Set OPENAI_API_KEY and/or ANTHROPIC_API_KEY as environment variables, or pass them directly:

from oaf.llmclient import LLMClient

client = LLMClient(openai_api_key="sk-...", anthropic_api_key="sk-ant-...")

License

MIT