A
licenseNot graded
qualityB
maintenanceProvides long-term memory for AI agents via MCP tools to store, recall, and delete memories, with per-user scoping and usage limits.
AGPL 3.0
Simple Central Agent Memory
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Simple Central Agent MemoryRemember that this project uses FalkorDB for cross-session memory."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Simple Central Agent Memory (SCAM)
A project-scoped memory gateway for AI agents, built with FastMCP, Graphiti,
and FalkorDB. The scam command starts the gateway; the existing
memory_gateway import and MCP tool names remain unchanged.
Installation
Requires Python 3.12 or newer. Install a locally built wheel with pip install path/to/simple_central_agent_memory-0.1.0-py3-none-any.whl, then run
scam --help. A published PyPI release is not yet available.
Running the service also requires a separately configured FalkorDB instance,
a compatible language model, an embedding model, and a reranker. Set the
required CENTRAL_MEMORY_*, EVREN_*, FALKORDB_URL, and embedding
environment variables before starting scam. Do not commit credentials.
Related MCP server: neo-mem
Safety and limitations
The current ingestion path schedules extraction in process: a queued
response is not a durable write, and a restart may lose pending work. Back up
the existing database and verify restore before migrating any deployment.
The current web UI's token handling and network transport require a security
review before public deployment. Do not expose this service to untrusted
networks without additional authentication and TLS.
This initial package is a staging copy. It does not replace or migrate an
existing deployment or its data. Licensed under the WTFPL; see LICENSE.
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo issues
Related MCP Connectors
Persistent memory, knowledge graph, task management and more for AI agents via MCP.
Cross-tool persistent memory and context for AI assistants over MCP.
- KogniteOAuthdev.kognite
Hosted agent memory: store, search, and recall facts across sessions from any MCP client.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceProvides persistent, graph-based memory for AI agents using Neo4j and vector embeddings, enabling semantic recall across sessions via MCP.1Apache 2.0
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to maintain and query project memory independently of the underlying model, with versioned, auditable storage and multi-stage retrieval through a single MCP gateway.MIT
- AlicenseNot gradedqualityCmaintenanceProvides persistent, multi-project memory for AI agents via MCP, enabling them to store and retrieve tasks, decisions, errors, checkpoints, and documentation while keeping context isolated by project.2MIT