Contextwell FAQs

Question 1

What is Contextwell and what does it do?

Accepted Answer

Contextwell is a persistent semantic memory server for AI assistants like GitHub Copilot CLI and VS Code. It allows you to store and recall various types of information—facts, decisions, code snippets, notes—across different sessions and projects using advanced semantic and hybrid search.

Question 2

Can I store different types of memories or make them project-specific?

Accepted Answer

Yes, Contextwell allows you to store various memory types including facts, decisions, code snippets, todos, and chat extracts. Memories can be scoped as 'global' (accessible everywhere) or 'project' (tied to your current Git repository), providing flexible organization.

Question 3

How does Contextwell enable AI assistants to 'remember' information?

Accepted Answer

Contextwell stores your memories in a LanceDB vector store. It uses embedding models to convert content into vector representations, enabling recall by meaning through semantic similarity search. For enhanced accuracy, it also supports hybrid retrieval with Reciprocal Rank Fusion (RRF).

Question 4

How can I add and retrieve memories with Contextwell?

Accepted Answer

You interact with Contextwell using CLI tools like `remember` to store new information and `recall` to search for memories by meaning. You can also ingest entire markdown files with `remember_file` and manage existing memories with `update` or `forget`.

Question 5

Which AI assistants and development environments are supported?

Accepted Answer

Contextwell is primarily designed to integrate with GitHub Copilot CLI and VS Code, functioning as an MCP (Memory, Context, and Preferences) server. This allows for seamless interaction and memory recall directly within your coding workflows.

Contextwell

Contextwell

主な機能

ユースケース

主な機能

ユースケース