knowledge_protocol.py
"""
Knowledge Protocol: Custom Knowledge Sources
==============================================
KnowledgeProtocol is an interface for building custom knowledge sources
that don't use the standard Knowledge class.
Implement this when you need:
- Knowledge from a non-standard source (file system, API, database)
- Custom search logic that doesn't fit the vector DB model
- Integration with existing retrieval systems
The protocol requires implementing build_context(), get_tools(), and aget_tools().
Optionally implement retrieve()/aretrieve() for the search_knowledge feature.
"""
from typing import Callable, List
from agno.agent import Agent
from agno.knowledge.document import Document
from agno.knowledge.protocol import KnowledgeProtocol
from agno.models.openai import OpenAIResponses
# ---------------------------------------------------------------------------
# Custom Knowledge Implementation
# ---------------------------------------------------------------------------
class InMemoryKnowledge(KnowledgeProtocol):
"""A simple in-memory knowledge source for demonstration.
In production, this could wrap a SQL database, REST API,
or any custom data source.
"""
def __init__(self):
self.documents: list[Document] = []
def add(self, name: str, content: str) -> None:
self.documents.append(Document(name=name, content=content))
def _search(self, query: str, limit: int = 5) -> List[Document]:
"""Simple substring matching (replace with your search logic)."""
results = []
for doc in self.documents:
if doc.content and query.lower() in doc.content.lower():
results.append(doc)
return results[:limit] or self.documents[:limit]
# --- Required protocol methods ---
def build_context(self, **kwargs) -> str:
return "Use the search tool to find information in the knowledge base."
def get_tools(self, **kwargs) -> List[Callable]:
return []
async def aget_tools(self, **kwargs) -> List[Callable]:
return []
# --- Optional: enables search_knowledge feature ---
def retrieve(self, query: str, **kwargs) -> List[Document]:
max_results = kwargs.get("max_results", 5)
return self._search(query, limit=max_results)
async def aretrieve(self, query: str, **kwargs) -> List[Document]:
return self.retrieve(query, **kwargs)
# ---------------------------------------------------------------------------
# Setup
# ---------------------------------------------------------------------------
custom_knowledge = InMemoryKnowledge()
custom_knowledge.add("Python", "Python is a high-level programming language.")
custom_knowledge.add("TypeScript", "TypeScript adds static types to JavaScript.")
custom_knowledge.add(
"Rust", "Rust is a systems language focused on safety and performance."
)
agent = Agent(
model=OpenAIResponses(id="gpt-5.2"),
knowledge=custom_knowledge,
search_knowledge=True,
markdown=True,
)
# ---------------------------------------------------------------------------
# Run Demo
# ---------------------------------------------------------------------------
if __name__ == "__main__":
print("\n" + "=" * 60)
print("Custom KnowledgeProtocol implementation")
print("=" * 60 + "\n")
agent.print_response("Tell me about Python", stream=True)
Run the Example
Set up your virtual environment
uv venv --python 3.12
source .venv/bin/activate
uv venv --python 3.12
.venv\Scripts\activate
Export your OpenAI API key
export OPENAI_API_KEY="your_openai_api_key_here"
$Env:OPENAI_API_KEY="your_openai_api_key_here"