LangGraph - Tools¶

This section documents all tools available to LangGraph agents.

Overview¶

Tools are decorated with @tool from LangChain and provide specific functionalities to agents. Each tool accepts a config parameter containing authentication and context information.

Available Tools¶

• User Tools - User profile retrieval
• Objective Tools - Career objectives management
• Opportunity Tools - Opportunity search and suggestions
• Task Helper Tools - AI assistant for professional questions
• Talent Finder Tools - Internal talent discovery
• Person Disambiguation Tools - Ambiguous person resolution

Tool Configuration¶

All tools accept a config parameter with the following structure:

{
    "configurable": {
        "user_application_id": "user123",  # Current user ID
        "client_id": "725",  # Client ID
        "data_source": "pwc",  # e.g. pwc, smartmobility
        "environment": "prod",  # Environment
    }
}

Usage Pattern¶

from src.chatbot.agentic.tools.user_tools import get_user_tool

# Call tool with config
result = await get_user_tool.ainvoke(
    {"user_id": "user456"},
    config={
        "configurable": {
            "user_application_id": "current_user",
            "client_id": "725",
            "data_source": "pwc",
        }
    },
)

Tool Development¶

When creating new tools:

1. Use @tool decorator from LangChain
2. Add Google-style docstrings
3. Include clear parameter descriptions
4. Handle errors gracefully
5. Log important actions
6. Return structured data