pip install hystersis openai

from openai import OpenAI
from hystersis import Hystersis

memory = Hystersis(api_key="your-hystersis-key")
llm = OpenAI(api_key="your-openai-key")

def chat_with_memory(user_id: str, message: str) -> str:
    # 1. Retrieve relevant memories
    memories = memory.search(
        query=message,
        user_id=user_id,
        limit=10
    )

    # 2. Build system prompt with memory context
    memory_context = "\n".join([
        f"- {m['content']}" for m in memories
    ])

    system_prompt = f"""You are a helpful assistant with long-term memory.
You remember facts about the user across conversations.
Relevant context from memory:
{memory_context}"""

    # 3. Send to OpenAI
    response = llm.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": system_prompt},
            {"role": "user", "content": message}
        ]
    )

    # 4. Store the conversation as memory
    memory.create_memory(
        content=f"User: {message} | Assistant: {response.choices[0].message.content}",
        user_id=user_id,
        compression_mode="extract"
    )

    return response.choices[0].message.content

# Usage
result = chat_with_memory("user_123", "I prefer dark mode for my IDE")
print(result)

def stream_with_memory(user_id: str, message: str):
    memories = memory.search(query=message, user_id=user_id, limit=5)
    memory_context = "\n".join([f"- {m['content']}" for m in memories])

    stream = llm.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": f"You have memory. Context:\n{memory_context}"},
            {"role": "user", "content": message}
        ],
        stream=True
    )

    full_response = ""
    for chunk in stream:
        if chunk.choices[0].delta.content:
            print(chunk.choices[0].delta.content, end="", flush=True)
            full_response += chunk.choices[0].delta.content

    # Store after streaming completes
    memory.create_memory(
        content=f"User: {message} | Assistant: {full_response}",
        user_id=user_id,
        compression_mode="extract"
    )

    return full_response

def chat_with_tools(user_id: str, message: str):
    memories = memory.search(query=message, user_id=user_id, limit=5)
    memory_context = "\n".join([f"- {m['content']}" for m in memories])

    tools = [
        {
            "type": "function",
            "function": {
                "name": "save_memory",
                "description": "Save important information to long-term memory",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "content": {"type": "string", "description": "Information to remember"},
                        "importance": {"type": "string", "enum": ["high", "medium", "low"]}
                    },
                    "required": ["content"]
                }
            }
        },
        {
            "type": "function",
            "function": {
                "name": "search_memory",
                "description": "Search long-term memory for information",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "query": {"type": "string", "description": "Search query"},
                        "limit": {"type": "integer", "description": "Max results", "default": 5}
                    },
                    "required": ["query"]
                }
            }
        }
    ]

    response = llm.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": f"You have memory tools. Context:\n{memory_context}"},
            {"role": "user", "content": message}
        ],
        tools=tools
    )

    # Process tool calls
    message = response.choices[0].message
    if message.tool_calls:
        for tool_call in message.tool_calls:
            if tool_call.function.name == "save_memory":
                result = memory.create_memory(
                    content=eval(tool_call.function.arguments)["content"],
                    user_id=user_id
                )
            elif tool_call.function.name == "search_memory":
                args = eval(tool_call.function.arguments)
                result = memory.search(
                    query=args["query"],
                    user_id=user_id,
                    limit=args.get("limit", 5)
                )
    return message

def create_memory_assistant(user_id: str):
    # Create an assistant with memory instructions
    assistant = llm.beta.assistants.create(
        name="Memory Assistant",
        instructions="""You are a helpful assistant with long-term memory.
Always check memory before responding.
Store important facts about the user.""",
        model="gpt-4o"
    )

    # Create a thread for this user
    thread = llm.beta.threads.create(
        metadata={"user_id": user_id}
    )

    return assistant, thread

def chat_assistant(user_id: str, message: str):
    assistant, thread = create_memory_assistant(user_id)

    # Retrieve memories
    memories = memory.search(query=message, user_id=user_id, limit=5)

    # Add message to thread
    llm.beta.threads.messages.create(
        thread_id=thread.id,
        role="user",
        content=f"[Memory Context]: {'; '.join([m['content'] for m in memories])}\n\n[User Message]: {message}"
    )

    # Run assistant
    run = llm.beta.threads.runs.create_and_poll(
        thread_id=thread.id,
        assistant_id=assistant.id
    )

    # Get response
    messages = llm.beta.threads.messages.list(thread_id=thread.id)
    response = messages.data[0].content[0].text.value

    # Store in memory
    memory.create_memory(
        content=f"User: {message} | Assistant: {response}",
        user_id=user_id
    )

    return response

# .env
COMPRESSION_LLM_FAST_PROVIDER=openai
COMPRESSION_LLM_FAST_MODEL=gpt-4o-mini