nbagents 0.1.1

A nano, minimalistic, and lightweight library for building ReACT-based AI agents that use tools to solve tasks.

🚀 Nbagents Summary

What We Added

✨ Async Support

• Full async/await support for both tools and LLM calls
• run_async() method for async execution
• is_async_function() utility to detect async functions
• run_sync_or_async() utility to handle both sync and async functions seamlessly

🔧 Built-in Tools

• Web Search: Search the web using DuckDuckGo API (both sync and async versions)
• File Operations:
  • read_file: Read content from files
  • write_file: Write content to files
  • list_files: List files in directories
• Time: get_current_time - Get current date and time
• Auto-registration of built-in tools (can be disabled with include_builtin_tools=False)

📋 Comprehensive Logging

• setup_logger() function for easy logger configuration
• Agent-specific loggers with configurable log levels
• Debug logging for tool execution and iterations
• Info/Warning/Error logging for key events
• Tool registration and execution logging

🔄 Enhanced Error Handling

• Better retry mechanisms for both LLM calls and tool execution
• Graceful error reporting with detailed messages
• Tool execution attempt logging
• Improved exception handling

🛠️ Improved Tool System

• Enhanced Tool class with async support detection
• Better tool schema parsing
• Tool metadata including async status
• Improved tool registry with built-in tool management

🎯 Better Agent Configuration

• log_level parameter for controlling logging verbosity
• include_builtin_tools parameter to control built-in tool loading
• Enhanced initialization with better defaults

📦 Dependencies Added

• requests - For web search functionality
• aiohttp - For async web search
• asyncio - For async support (built-in)
• logging - For comprehensive logging (built-in)

🎮 Usage Examples

Basic Usage with Built-in Tools

from nbagents import Agent

agent = Agent(llm=your_llm_function, include_builtin_tools=True)
result = agent.run("Get the current time and save it to a file")

Async Usage

import asyncio
from nbagents import Agent

async def main():
    agent = Agent(llm=async_llm_function, include_builtin_tools=True)
    result = await agent.run_async("Search for Python tutorials and save results")

asyncio.run(main())

Custom Logging

from nbagents import Agent, setup_logger

logger = setup_logger("my_agent", "DEBUG")
agent = Agent(llm=your_llm, log_level="DEBUG")

Custom Tools with Async

@agent.tool("async_process")
async def async_process(data: str) -> str:
    await asyncio.sleep(1)  # Simulate async work
    return f"Processed: {data}"

✅ Verified Working Features

• ✅ File read/write operations work correctly
• ✅ Time retrieval works correctly
• ✅ Async tool execution works
• ✅ Comprehensive logging is functional
• ✅ Error handling and retries work
• ✅ Tool registration and discovery works
• ✅ Both sync and async LLM calls supported

🚀 Next Possible Enhancements

• Memory/persistence system
• Multi-agent communication
• Plugin system
• Configuration file support
• Performance metrics and monitoring
• Additional built-in tools (database, API calls, etc.)
• Tool composition and chaining
• Custom prompt templates