my-project-public

repository

loading code, commits, and activity

repositories

loading repo index

#1	---
#2	title: Agno
#3	---
#4
#5	This integration of [Mem0](https://github.com/mem0ai/mem0) with [Agno](https://github.com/agno-agi/agno) enables persistent, multimodal memory for Agno-based agents - improving personalization, context awareness, and continuity across conversations.
#6
#7	## Overview
#8
#9	1. Store and retrieve memories from Mem0 within Agno agents
#10	2. Support for multimodal interactions (text and images)
#11	3. Semantic search for relevant past conversations
#12	4. Personalized responses based on user history
#13	5. One-line memory integration via `Mem0Tools`
#14
#15	## Prerequisites
#16
#17	Before setting up Mem0 with Agno, ensure you have:
#18
#19	1. Installed the required packages:
#20	```bash
#21	pip install agno mem0ai python-dotenv
#22	```
#23
#24	2. Valid API keys:
#25	- [Mem0 API Key](https://app.mem0.ai/dashboard/api-keys)
#26	- OpenAI API Key (for the agent model)
#27
#28	## Quick Integration (Using `Mem0Tools`)
#29
#30	The simplest way to integrate Mem0 with Agno Agents is to use Mem0 as a tool using built-in `Mem0Tools`:
#31
#32	```python
#33	from agno.agent import Agent
#34	from agno.models.openai import OpenAIChat
#35	from agno.tools.mem0 import Mem0Tools
#36
#37	agent = Agent(
#38	name="Memory Agent",
#39	model=OpenAIChat(id="gpt-4.1-nano-2025-04-14"),
#40	tools=[Mem0Tools()],
#41	description="An assistant that remembers and personalizes using Mem0 memory."
#42	)
#43	```
#44
#45	This enables memory functionality out of the box:
#46
#47	- Persistent memory writing: `Mem0Tools` uses `MemoryClient.add(...)` to store messages from user-agent interactions, including optional metadata such as user ID or session.
#48	- Contextual memory search: Compatible queries use `MemoryClient.search(...)` to retrieve relevant past messages, improving contextual understanding.
#49	- Multimodal support: Both text and image inputs are supported, allowing richer memory records.
#50
#51	> `Mem0Tools` uses the `MemoryClient` under the hood and requires no additional setup. You can customize its behavior by modifying your tools list or extending it in code.
#52
#53	## Full Manual Example
#54
#55	> Note: Mem0 can also be used with Agno Agents as a separate memory layer.
#56
#57	The following example demonstrates how to create an Agno agent with Mem0 memory integration, including support for image processing:
#58
#59	```python
#60	import base64
#61	from pathlib import Path
#62	from typing import Optional
#63
#64	from agno.agent import Agent
#65	from agno.media import Image
#66	from agno.models.openai import OpenAIChat
#67	from mem0 import MemoryClient
#68
#69	# Initialize the Mem0 client
#70	client = MemoryClient()
#71
#72	# Define the agent
#73	agent = Agent(
#74	name="Personal Agent",
#75	model=OpenAIChat(id="gpt-4"),
#76	description="You are a helpful personal agent that helps me with day to day activities."
#77	"You can process both text and images.",
#78	markdown=True
#79	)
#80
#81
#82	def chat_user(
#83	user_input: Optional[str] = None,
#84	user_id: str = "alex",
#85	image_path: Optional[str] = None
#86	) -> str:
#87	"""
#88	Handle user input with memory integration, supporting both text and images.
#89
#90	Args:
#91	user_input: The user's text input
#92	user_id: Unique identifier for the user
#93	image_path: Path to an image file if provided
#94
#95	Returns:
#96	The agent's response as a string
#97	"""
#98	if image_path:
#99	# Convert image to base64
#100	with open(image_path, "rb") as image_file:
#101	base64_image = base64.b64encode(image_file.read()).decode("utf-8")
#102
#103	# Create message objects for text and image
#104	messages = []
#105
#106	if user_input:
#107	messages.append({
#108	"role": "user",
#109	"content": user_input
#110	})
#111
#112	messages.append({
#113	"role": "user",
#114	"content": {
#115	"type": "image_url",
#116	"image_url": {
#117	"url": f"data:image/jpeg;base64,{base64_image}"
#118	}
#119	}
#120	})
#121
#122	# Store messages in memory
#123	client.add(messages, user_id=user_id)
#124	print("✅ Image and text stored in memory.")
#125
#126	if user_input:
#127	# Search for relevant memories
#128	memories = client.search(user_input, user_id=user_id)
#129	memory_context = "\n".join(f"- {m['memory']}" for m in memories['results'])
#130
#131	# Construct the prompt
#132	prompt = f"""
#133	You are a helpful personal assistant who helps users with their day-to-day activities and keeps track of everything.
#134
#135	Your task is to:
#136	1. Analyze the given image (if present) and extract meaningful details to answer the user's question.
#137	2. Use your past memory of the user to personalize your answer.
#138	3. Combine the image content and memory to generate a helpful, context-aware response.
#139
#140	Here is what I remember about the user:
#141	{memory_context}
#142
#143	User question:
#144	{user_input}
#145	"""
#146	# Get response from agent
#147	if image_path:
#148	response = agent.run(prompt, images=[Image(filepath=Path(image_path))])
#149	else:
#150	response = agent.run(prompt)
#151
#152	# Store the interaction in memory
#153	interaction_message = [{"role": "user", "content": f"User: {user_input}\nAssistant: {response.content}"}]
#154	client.add(interaction_message, user_id=user_id)
#155	return response.content
#156
#157	return "No user input or image provided."
#158
#159
#160	# Example Usage
#161	if __name__ == "__main__":
#162	response = chat_user(
#163	"I like to travel and my favorite destination is London",
#164	image_path="travel_items.jpeg",
#165	user_id="alex"
#166	)
#167	print(response)
#168	```
#169
#170	## Key Features
#171
#172	### 1. Multimodal Memory Storage
#173
#174	The integration supports storing both text and image data:
#175
#176	- Text Storage: Conversation history is saved in a structured format
#177	- Image Analysis: Agents can analyze images and store visual information
#178	- Combined Context: Memory retrieval combines both text and visual data
#179
#180	### 2. Personalized Agent Responses
#181
#182	Improve your agent's context awareness:
#183
#184	- Memory Retrieval: Semantic search finds relevant past interactions
#185	- User Preferences: Personalize responses based on stored user information
#186	- Continuity: Maintain conversation threads across multiple sessions
#187
#188	### 3. Flexible Configuration
#189
#190	Customize the integration to your needs:
#191
#192	- Use `Mem0Tools()` for drop-in memory support
#193	- Use `MemoryClient` directly for advanced control
#194	- User Identification: Organize memories by user ID
#195	- Memory Search: Configure search relevance and result count
#196	- Memory Formatting: Support for various OpenAI message formats
#197
#198	<CardGroup cols={2}>
#199	<Card title="OpenAI Agents SDK" icon="cube" href="/integrations/openai-agents-sdk">
#200	Build agents with OpenAI SDK and Mem0
#201	</Card>
#202	<Card title="Mastra Integration" icon="star" href="/integrations/mastra">
#203	Create intelligent agents with Mastra framework
#204	</Card>
#205	</CardGroup>
#206
#207

z6Mkq5mY3JWtxoxUobWcfNHm7AkRubgSWEZTkBVqZXJviFZ5/my-project-public