EngramMemoryEngramMemoryOfficialยท1 Agent Skills Included

engram-memory

Persistent cross-session memory with fast semantic recall

Stores, searches, recalls, and forgets long-term memories across sessions using local semantic embeddings. Eliminates re-explaining your codebase, preferences, and past decisions every time a new session starts. Runs fully self-hosted in one Docker container with a three-tier recall engine that gets faster the more you use it. Works with Claude Code, Cursor, Windsurf, VS Code, and OpenClaw via MCP.
npx skills add EngramMemory/engram-memory --all -g -y

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install Engram Memory?โ–ผ

Run `npx skills add EngramMemory/engram-memory --all -g -y` in your terminal to install the skill globally, then start the bundled Docker container to enable local storage and embeddings.

How to give an AI agent persistent memory?โ–ผ

Engram stores every important fact, preference, and decision as a semantic embedding, then auto-recalls relevant memories at the start of each new session so nothing is forgotten.

Does Engram send my data to the cloud?โ–ผ

No. The community edition runs entirely on your own hardware with local embeddings and a local vector database, so no API keys or external servers are required.

Which AI clients work with Engram?โ–ผ

It works with Claude Code, Cursor, Windsurf, VS Code, OpenClaw, and any other client that supports the MCP protocol.

How fast is Engram memory recall?โ–ผ

Frequently used memories return in about 25ms from the hot-tier cache, while full semantic searches complete in roughly 190ms, getting faster the more you use it.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’