engram-memory-community

<div align="center">

Three-Tiered Brain for AI agents. Self-hosted. Zero API costs.

Docs · Quickstart · Dashboard · Cloud SDKs

</div>

---

Engram gives your AI agent persistent memory across sessions. Store, search, recall, and forget memories using semantic embeddings — all running on your own hardware with Qdrant and FastEmbed.

One repo, two interfaces: an OpenClaw skill and a universal MCP server that works with Claude Code, Cursor, Windsurf, and VS Code.

---

| Tool | What it does | |---|---| | memory_store | Save a memory with semantic embedding and auto-classification | | memory_search | Semantic similarity search across all stored memories | | memory_recall | Auto-inject relevant memories into agent context | | memory_forget | Remove memories by ID or search match |

Categories: preference, fact, decision, entity, other — auto-detected from content.

---

# Requires Docker
bash scripts/setup.sh

This starts Qdrant (vector DB) and FastEmbed (local embedding model) on your machine.

OpenClaw (replaces default memory system):

# Clone and install the plugin
git clone https://github.com/EngramMemory/engram-memory-community.git
cd engram-memory-community
bash scripts/install-plugin.sh

This installs Engram as a plugin (not a skill) and sets it as the memory backend, replacing the built-in SQLite memory with the three