langchain-short-memory

What problem does it solve?

Implement and tune short-term memory for LangChain/LangGraph chat agents, including thread-level persistence, trimming, deletion, and summarization patterns when conversation history grows.

Core Features & Use Cases

• Thread-scoped memory (thread_id) for multi-turn conversations.
• Memory lifecycle patterns: append, trim, summarize, selective delete.
• Practical defaults that avoid context bloat.

Quick Start

Instantiate a LangChain agent with thread-scoped memory and enable trimming, summarization, and selective deletion to manage conversation history.