57 lines
1.6 KiB
Python
57 lines
1.6 KiB
Python
"""Agent tools for interacting with the knowledge base.
|
||
|
||
This module defines two tools that can be used by the LangChain agent:
|
||
|
||
* ``search_knowledge_base`` – performs a semantic search in the Qdrant vector store.
|
||
* ``add_to_knowledge_base`` – adds a new document (title + content) to the store.
|
||
|
||
Both tools are decorated with ``@tool`` from ``langchain.tools`` so that they can be
|
||
exposed to the agent.
|
||
"""
|
||
|
||
from typing import List, Dict, Any
|
||
|
||
from langchain.tools import tool
|
||
|
||
from .vector_store import KnowledgeBase
|
||
|
||
# Create a single global knowledge base instance that all tools will use.
|
||
# In a real deployment you might want to inject this via dependency injection.
|
||
kb = KnowledgeBase()
|
||
|
||
@tool("search_knowledge_base")
|
||
def search_knowledge_base(query: str, max_results: int = 5) -> List[Dict[str, Any]]:
|
||
"""Search the knowledge base for relevant chunks.
|
||
|
||
Parameters
|
||
----------
|
||
query: str
|
||
The search query.
|
||
max_results: int, optional
|
||
Number of top results to return. Defaults to 5.
|
||
|
||
Returns
|
||
-------
|
||
List[Dict[str, Any]]
|
||
A list of dictionaries containing ``content``, ``title``, ``chunk_index`` and ``score``.
|
||
"""
|
||
return kb.search(query, max_results)
|
||
|
||
@tool("add_to_knowledge_base")
|
||
def add_to_knowledge_base(content: str, title: str) -> str:
|
||
"""Add a new document to the knowledge base.
|
||
|
||
Parameters
|
||
----------
|
||
content: str
|
||
Full text of the document.
|
||
title: str
|
||
Title or name of the document.
|
||
|
||
Returns
|
||
-------
|
||
str
|
||
Confirmation message.
|
||
"""
|
||
kb.add_document(content, title)
|
||
return f"Document '{title}' added to the knowledge base." |