org.slashlib.py.agent 0.1.0

python agent for inference handling

Bottom AI CHANGELOG LICENSE

org.slashlib.py.agent

A highly decoupled, asynchronous framework for building AI agents in Python.

Core Concept

This package provides a robust infrastructure to connect AI models (Inference Engines) with functional tools. The focus lies on Provider Agnosticism: The agent does not need to know whether it is communicating with Ollama, OpenAI, or a local model—it uses standardized adapters to ensure seamless integration.

Key Features

• Asynchronous Core: Built on asyncio for non-blocking task execution.
• Provider Agnostic: Easily swap the AI engine using the Adapter pattern.
• Automatic Tool Schemas: Automatically transforms Python functions into JSON schemas for LLMs via decorators.
• Multiton Pattern: Ensures unique agent instances by identifier, preventing redundant resource allocation.
• Robust Exception Hierarchy: Clearly separates connection, configuration, and tool execution errors.

---

Installation

Install the package via pip:

pip install org.slashlib.py.agent

---

Quick Start

Setting up an agent with a tool and the Ollama adapter is straightforward:

import asyncio
from org.slashlib.py.agent import Agent, OllamaInferenceAdapter, tool

# 1. Define a tool
@tool(description="Adds two numbers.")
async def add_numbers(a: int, b: int) -> int:
    return a + b

async def main():
    # 2. Configure Adapter and Agent
    adapter = OllamaInferenceAdapter()
    my_agent = Agent(
        identifier="MathExpert",
        tools=[add_numbers],
        adapter=adapter
    )

    # 3. Start task (non-blocking)
    task = my_agent.run(user_prompt="What is 123 + 456?")
    
    # 4. Retrieve result
    response = await task
    print(f"Response: {response.get_last_content()}")

if __name__ == "__main__":
    asyncio.run(main())

---

Documentation & Obsidian

The project root is pre-configured as an Obsidian Vault. If you open this folder directly in Obsidian, all settings and documentation links will be available immediately via the included .obsidian directory.

The following community plugins are pre-configured in the vault to enhance the documentation experience:

• File Include: Embed code files directly into your markdown documentation.
• Folder Notes: Add descriptions at the folder level.
• Front Matter Title: Use metadata for descriptive file titles.
• Hide Folders: Keeps the structure clean by hiding internal directories.
• Iconic & Icons: Improved visual navigation.

More docs

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

© 2026 org.slashlib

TOP AI CHANGELOG LICENSE