basic-open-agent-tools

An open foundational toolkit providing essential components for building AI agents with minimal dependencies for local (non-HTTP/API) actions. Designed with agent-friendly type signatures to eliminate "signature too complex" errors, while offering core utilities that developers can easily integrate into their agents to avoid excess boilerplate.

Installation

pip install basic-open-agent-tools

Or with UV:

uv add basic-open-agent-tools

Key Features

✨ Agent-Friendly Design: All functions use simplified type signatures to prevent "signature too complex" errors when used with AI agent frameworks

🚀 Minimal Dependencies: Pure Python implementation with no external dependencies for core functionality

🔧 Modular Architecture: Load only the tools you need with category-specific helpers

Quick Start

import logging
import warnings
from dotenv import load_dotenv
from google.adk.agents import Agent
from google.adk.models.lite_llm import LiteLlm

import basic_open_agent_tools as boat

# Load tools by category
fs_tools = boat.load_all_filesystem_tools()    # 18 functions
text_tools = boat.load_all_text_tools()       # 10 functions

# Merge for agent use (automatically deduplicates)
agent_tools = boat.merge_tool_lists(fs_tools, text_tools)


load_dotenv()

agent_instruction = """
**INSTRUCTION:**
You are FileOps, a specialized file and directory operations sub-agent.
Your role is to execute file operations (create, read, update, delete, move, copy) and directory operations (create, delete) with precision.
**Guidelines:**
- **Preserve Content:** Always read full file content before modifications; retain all original content except targeted changes.
- **Precision:** Execute instructions exactly, verify operations, and handle errors with specific details.
- **Communication:** Provide concise, technical status reports (success/failure, file paths, operation type, content preservation details).
- **Scope:** File/directory CRUD, move, copy, path validation. No code analysis.
- **Confirmation:** Confirm completion to the senior developer with specific details of modifications.
"""

logging.basicConfig(level=logging.ERROR)
warnings.filterwarnings("ignore")

file_ops_agent = Agent(
    model=LiteLlm(model="anthropic/claude-3-5-haiku-20241022"),
    name="FileOps",
    instruction=agent_instruction,
    description="Specialized file and directory operations sub-agent for the Python developer.",
    tools=agent_tools,
)

"""
The above would load:

File and Directory Operations:
    read_file_to_string
    write_file_from_string
    append_to_file
    list_directory_contents
    create_directory
    delete_file
    delete_directory
    move_file
    copy_file
    get_file_info
    file_exists
    directory_exists
    get_file_size
    is_empty_directory
    list_all_directory_contents
    generate_directory_tree
    validate_path
    validate_file_content

Text Processing Tools:
    clean_whitespace
    normalize_line_endings
    strip_html_tags
    normalize_unicode
    to_snake_case
    to_camel_case
    to_title_case
    smart_split_lines
    extract_sentences
    join_with_oxford_comma

"""

Documentation

• Getting Started - Installation and quick start guide
• Examples - Detailed usage examples and patterns
• Contributing - Development setup and guidelines

Current Features

File System Tools ✅ (18 functions)

• File operations (read, write, append, delete, copy, move)
• Directory operations (create, list, delete, tree visualization)
• File information and existence checking
• Path validation and error handling

Text Processing Tools ✅ (10 functions)

• Text cleaning and normalization
• Case conversion utilities (snake_case, camelCase, Title Case)
• Smart text splitting and sentence extraction
• HTML tag removal and Unicode normalization

Data Tools ✅ (28+ functions with Agent-Friendly Signatures)

Enhanced for AI Agents: All functions use simplified type signatures for maximum compatibility

• Structure Tools: flatten_dict_simple, merge_dicts_simple, get_nested_value_simple
• CSV Tools: read_csv_simple, write_csv_simple with basic types
• Validation Tools: validate_schema_simple, validate_data_types_simple
• JSON & Config: Safe serialization with compression and validation
• Transform Tools: Data cleaning and normalization functions

Additional Data Modules 📋:

• Object serialization and configuration files (YAML, TOML, INI)
• Binary data processing and archive handling
• Advanced transformation and streaming utilities

✨ Agent Compatibility: All data functions use basic Python types (str, dict, list, bool) instead of complex Union types or custom type aliases, ensuring seamless integration with AI agent frameworks.

Future Modules 🚧

• Network Tools - HTTP utilities, API helpers
• System Tools - Process management, system information
• Crypto Tools - Hashing, encoding, basic cryptographic utilities
• Utilities - Development and debugging helpers

Contributing

We welcome contributions! Please see our Contributing Guide for development setup, coding standards, and pull request process.