
suma-memory
MCP server by github.com/sumanaddanki/suma-mcp
Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
About suma-memory
suma-memory is an MCP (Model Context Protocol) server published by sumanaddanki in the official MCP registry, listed under Knowledge & Memory on Loomal. Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
suma-memory runs as a hosted remote over streamable-http — MCP clients connect directly to its endpoint, with nothing to install locally. It also ships as an npm package (suma-mcp-proxy), so any MCP client that can launch a local process can run it.
Development happens in the open at github.com/sumanaddanki/suma-mcp.
Use suma-memory with your agent
claude mcp add --transport http suma-memory https://sumapro.quadframe.work/mcp{
"mcpServers": {
"suma-memory": {
"url": "https://sumapro.quadframe.work/mcp"
}
}
}https://sumapro.quadframe.work/mcpsuma-mcp-proxyFrequently asked questions
- What is suma-memory?
- suma-memory is an MCP (Model Context Protocol) server by sumanaddanki in the Knowledge & Memory category. Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
- How do I connect suma-memory to Claude, Cursor, or another MCP client?
- suma-memory is a remote MCP server — add its endpoint (https://sumapro.quadframe.work/mcp) to your client's MCP configuration (for example with "claude mcp add", or under "mcpServers" in Cursor's mcp.json) and restart the client.
- Is suma-memory open source?
- Yes — the source code is public at github.com/sumanaddanki/suma-mcp.
- Can AI agents pay to use suma-memory?
- Not yet through Loomal — suma-memory is listed as a free directory entry. If its maintainer verifies ownership, they can set per-call USDC pricing that agents pay over x402, with settlement on Base.
More Knowledge & Memory MCP servers

Codebase Memory
3.4k
Codebase knowledge graph for AI agents — 159 languages, sub-ms queries, 99% fewer tokens.

basic-memory
3.2k
Local-first knowledge management with bi-directional LLM sync via Markdown files.

tradememory-protocol
1.2k
MCP memory for AI trading agents. Store trades, recall similar setups, track strategy performance.

mcp-neo4j-memory
959
MCP Neo4j Knowledge Graph Memory Server

ori-memory
310
Persistent memory infrastructure for AI agents. Identity, knowledge graph, and search.

YourMemory
245
Persistent memory for AI agents. Memories decay with time and strengthen on recall.
Listing data from the official MCP registry and GitHub, refreshed periodically. Not affiliated with the maintainer unless claimed. Maintain suma-memory? Claim this listing free by verifying GitHub ownership, or contact us.