From 313d2c8e6ed5e2c2bc6a88f0fb41c71a8dc1e9d6 Mon Sep 17 00:00:00 2001
From: Morven <81895400+nullure@users.noreply.github.com>
Date: Sun, 7 Dec 2025 21:13:04 +0530
Subject: [PATCH] Create openmemry.mdx

---
 .../integrations/retrievers/openmemry.mdx     | 129 ++++++++++++++++++
 1 file changed, 129 insertions(+)
 create mode 100644 src/oss/python/integrations/retrievers/openmemry.mdx

diff --git a/src/oss/python/integrations/retrievers/openmemry.mdx b/src/oss/python/integrations/retrievers/openmemry.mdx
new file mode 100644
index 0000000000..5f6003fc27
--- /dev/null
+++ b/src/oss/python/integrations/retrievers/openmemry.mdx
@@ -0,0 +1,129 @@
+---
+title: OpenMemory
+---
+
+# OpenMemory
+
+[OpenMemory](https://github.com/CaviraOSS/OpenMemory) is a local-first, persistent temporal memory engine implemented entirely through a Python client. It provides storage and retrieval of conversational and factual memory without requiring vector databases or embedding pipelines.
+
+`OpenMemoryRetriever` is an integration module that returns documents from a temporal memory store. It is based on the `BaseRetriever` class and enables persistent recall across sessions using only the `openmemory-py` Python package.
+
+This guide will help you get started with the OpenMemory [retriever](/oss/langchain/retrieval). For detailed documentation of configuration options and storage behavior, see the OpenMemory repository.
+
+### Integration details
+
+<ItemTable category="document_retrievers" item="OpenMemoryRetriever" />
+
+## Setup
+
+To use this module, you need:
+
+- Python 3.9 or higher
+- The `openmemory-py` Python package
+- The `langchain-openmemory` integration package
+
+If you want to enable tracing, you can also set your LangSmith API key:
+
+```python
+import os
+
+os.environ["LANGSMITH_API_KEY"] = "your_api_key"
+os.environ["LANGSMITH_TRACING"] = "true"
+```
+
+## Installation
+
+This retriever lives in the `langchain-openmemory` package:
+
+```bash
+pip install openmemory-py langchain-openmemory
+```
+
+## Instantiation
+
+The `Memory` class exposes the retriever interface directly. No configuration is required by default.
+
+```python
+from langchain_openmemory import Memory
+
+memory = Memory()
+retriever = memory.retriever
+```
+
+Optional explicit user separation can be provided:
+
+```python
+memory = Memory(user="user123")
+retriever = memory.retriever
+```
+
+If the `user` parameter is omitted, OpenMemory manages session identity internally.
+
+## Usage
+
+Once instantiated, the retriever can be used to retrieve documents from persistent memory.
+
+```python
+retriever.invoke("Where does the user live?")
+```
+
+This returns a list of LangChain Document objects.
+
+## Example
+
+This section demonstrates basic usage of OpenMemory as a persistent memory retriever.
+
+```python
+from langchain_openmemory import Memory
+
+memory = Memory()
+
+memory.store("The user lives in Hyderabad.")
+docs = memory.retriever.invoke("Where does the user live?")
+
+for doc in docs:
+    print(doc.page_content)
+```
+
+## Use within a chain
+
+```python
+from langchain_core.output_parsers import StrOutputParser
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.runnables import RunnablePassthrough
+from langchain_openai import ChatOpenAI
+from langchain_openmemory import Memory
+
+memory = Memory()
+retriever = memory.retriever
+
+prompt = ChatPromptTemplate.from_template(
+    """Answer the question based only on the context provided.
+
+Context: {context}
+
+Question: {question}"""
+)
+
+llm = ChatOpenAI()
+
+
+def format_docs(docs):
+    return "\n\n".join(doc.page_content for doc in docs)
+
+
+chain = (
+    {"context": retriever | format_docs, "question": RunnablePassthrough()}
+    | prompt
+    | llm
+    | StrOutputParser()
+)
+
+chain.invoke("Where does the user live?")
+```
+
+## API reference
+
+For implementation details and configuration options, see the OpenMemory repository:
+
+https://github.com/CaviraOSS/OpenMemory