Memory FAQs

Question 1

How does Memory work?

Accepted Answer

Memory leverages PostgreSQL and the pgvector extension to store and retrieve memories based on semantic similarity. It automatically generates embeddings using BERT and allows you to search, create, and manage memories via a RESTful API.

Question 2

What kind of data can be stored in Memory?

Accepted Answer

Memory can store various types of data as memories, including learnings, experiences, and other relevant information. Each memory includes content, source, tags, a confidence score, and a vector embedding for semantic search.

Question 3

What are the key features of Memory?

Accepted Answer

Key features include PostgreSQL with pgvector for vector similarity search, automatic embedding generation, a RESTful API, semantic search, support for different memory types and tags, and real-time updates via Server-Sent Events (SSE).

Question 4

What is Memory?

Accepted Answer

Memory is a server that adds long-term memory capabilities to AI assistants. It uses PostgreSQL with pgvector for efficient vector similarity search and provides a RESTful API for memory operations.

Question 5

How do I integrate Memory with Cursor?

Accepted Answer

You can integrate Memory with Cursor by adding a configuration block to your `~/.cursor/mcp.json` file, specifying the path to your Memory server installation. Cursor will then automatically start the server when needed.

Memory

关于

主要功能

使用案例