​

Manual Knowledge Filters

​

Step 1: Attach Metadata

1. Attach Metadata When Initializing the Knowledge Base
   knowledge_base = PDFKnowledgeBase(
       path=[
           {
               "path": "path/to/cv1.pdf",
               "metadata": {
                   "user_id": "jordan_mitchell",
                   "document_type": "cv",
                   "year": 2025,
               },
           },
           # ... more documents ...
       ],
       vector_db=vector_db,
   )
   knowledge_base.load(recreate=True)
   
2. Attach Metadata When Loading Documents One by One
   # Initialize the PDFKnowledgeBase
   knowledge_base = PDFKnowledgeBase(
       vector_db=vector_db,
       num_documents=5,
   )
   
   # Load first document with user_1 metadata
   knowledge_base.load_document(
       path=path/to/cv1.pdf,
       metadata={"user_id": "jordan_mitchell", "document_type": "cv", "year": 2025},
       recreate=True,  # Set to True only for the first run, then set to False
   )
   
   # Load second document with user_2 metadata
   knowledge_base.load_document(
       path=path/to/cv2.pdf,
       metadata={"user_id": "taylor_brooks", "document_type": "cv", "year": 2025},
   )
   

💡 Tips:
• Use Option 1 if you have all your documents and metadata ready at once.
• Use Option 2 if you want to add documents incrementally or as they become available.

​

Step 2: Query with Filters

​

1. On the Agent (applies to all queries)

agent = Agent(
    knowledge=knowledge_base,
    search_knowledge=True,
    knowledge_filters={"user_id": "jordan_mitchell"},
)
agent.print_response(
    "Tell me about Jordan Mitchell's experience and skills",
    markdown=True,
)

​

2. On Each Query (overrides Agent filters for that run)

agent = Agent(
    knowledge=knowledge_base,
    search_knowledge=True,
)
agent.print_response(
    "Tell me about Jordan Mitchell's experience and skills",
    knowledge_filters={"user_id": "jordan_mitchell"},
    markdown=True,
)

​

Combining Multiple Filters

agent = Agent(
    knowledge=knowledge_base,
    search_knowledge=True,
    knowledge_filters={
        "user_id": "jordan_mitchell",
        "document_type": "cv",
        "year": 2025,
    }
)
agent.print_response(
    "Tell me about Jordan Mitchell's experience and skills",
    markdown=True,
)

​

Try It Yourself!

• Load documents with different metadata.
• Query with different filter combinations.
• Observe how the results change!

​

Developer Resources

• Manual filtering
• Manual filtering on load