add-backend-tool

majiayu000's avatarfrom majiayu000

Add a new tool to the backend OpenAI function calling system. Use when user mentions "new tool", "add tool", "backend function", "agent capability", or wants to extend what the AI agent can do.

⭐0stars🔀0forks📁View on GitHub🕐Updated Jan 5, 2026⬇️1downloads
[Tool Routing and Orchestration]

When & Why to Use This Skill

This Claude skill facilitates the seamless expansion of an AI agent's capabilities by automating the integration of new tools into a backend OpenAI function-calling system. It provides a structured framework for defining Python helper functions, generating the required JSON schemas for tool definitions, and implementing execution handlers within the application's message loop. By streamlining the process of adding custom backend functions, it enables developers to rapidly evolve their agents from simple chat interfaces into powerful, action-oriented assistants capable of interacting with various APIs and local systems.

Use Cases

  • System Capability Expansion: Quickly adding new backend functions when a user requests a feature currently outside the agent's native capability set, such as directory listing or file manipulation.
  • DevOps and Git Integration: Implementing specialized tools for git operations, terminal command execution, or cloud infrastructure management to allow the agent to handle technical workflows.
  • Data Acquisition and Scraping: Extending the agent's reach by adding web scraping capabilities or custom API connectors to fetch and process real-time data from external sources.
  • Custom Business Logic Integration: Bridging the gap between LLM reasoning and proprietary internal systems by wrapping complex backend processes into callable functions for the agent.
nameadd-backend-tool
descriptionAdd a new tool to the backend OpenAI function calling system. Use when user mentions "new tool", "add tool", "backend function", "agent capability", or wants to extend what the AI agent can do.

Add Backend Tool

Instructions

  1. Read backend/main.py to understand existing tool patterns:

    • Find the tools list with function definitions
    • Review helper functions (read_file, write_file, run_terminal_command, web_search)

  2. Create the helper function:

    def new_tool_name(param1: str, param2: int = 10) -> str:
    """Docstring explaining the tool."""
    try:
        # Implementation
        return result
    except Exception as e:
        return f"Error: {str(e)}"

  3. Add tool definition to the tools list:

    {
    "type": "function",
    "function": {
        "name": "new_tool_name",
        "description": "What this tool does and when to use it",
        "parameters": {
            "type": "object",
            "properties": {
                "param1": {"type": "string", "description": "..."},
                "param2": {"type": "integer", "description": "..."}
            },
            "required": ["param1"]
        }
    }
}

  4. Add tool invocation handler in the WebSocket message loop:

    elif func_name == "new_tool_name":
    result = new_tool_name(**args)

  5. Update mode restrictions if needed:

    • Agent mode: full access
    • Chat mode: add to allowed list only if read-only/safe

Examples

  • "Add a tool to list directory contents"
  • "Create a tool for git operations"
  • "Add web scraping capability"

Guardrails

  • Include proper error handling with try/except
  • Add timeout for any long-running operations
  • Consider security implications (Chat mode restrictions)
  • Never hardcode API keys or secrets
  • Document the tool's purpose in the function docstring