Question 1

What types of information does Dukememory store?

Accepted Answer

Dukememory stores various types of project-critical information, including goals, architectural decisions, operational constraints, build commands, known issues, task progress, and design notes, all as structured memory cards.

Question 2

What is Dukememory?

Accepted Answer

Dukememory is a local memory layer specifically designed for AI coding agents (like Codex, Claude, Cursor). It helps these agents retain crucial project context, such as decisions, constraints, and task state, across multiple interactions without needing to re-send long chat histories.

Question 3

Is Dukememory a cloud-based service?

Accepted Answer

No, Dukememory is local-first. It stores all project memory directly within your project's `.agent/memory.db` file. No cloud service is required, and even semantic recall via embeddings can be configured to use local models like Ollama.

Question 4

Does Dukememory offer a user interface?

Accepted Answer

Yes, Dukememory includes a web UI that allows you to inspect, edit, and audit your project's memory. You can use it to search facts, review inbox items, monitor usage, and check the health of autonomous maintenance operations.

Question 5

How does Dukememory improve AI coding workflows?

Accepted Answer

It provides agents with compact task briefs, pre-edit impact checks, and durable, structured knowledge stored as typed memory cards. This reduces token waste, prevents agents from forgetting important context, and lowers overall context costs, leading to more efficient and accurate coding.

Dukememory

Client Configuration

System Overview

Frequently Asked Questions

Related MCP Servers

0xArchive

0xGasless

0xGasless