Databases / Mnemosyne
Mnemosyne
io.github.28naem-del/mnemosyne · v3.0.0-rc.4
Evidence-tracked local memory MCP server: one SQLite file; corrections retire dependent memories.
Description supplied by the publisher in the MCP Registry. We have not tested this server.
Install
npx -y mnemosy-ai mcp --db /absolute/path/project/.mnemosyne/memory.sqlite --workspace my-project --agent assistantAssembled from the package metadata in the official MCP Registry (including any subcommand or flag the publisher lists). Some publishers do not list the subcommand that starts MCP mode, so confirm the command in the README, along with any required configuration and credentials, before connecting it to real data.
Signals
| GitHub stars | 155 |
| npm weekly downloads | 774 |
| Last push | 4d ago |
| Licence | MIT |
| Open issues | 8 |
| SECURITY.md | Yes |
| Language | TypeScript |
| Latest version published | 2026-10-29 |
Signals captured from GitHub, npm and the official MCP Registry on November 2, 2026. See the scoring method.
Similar servers compare all →
rmbr
Score 73Embedded, local-first memory and retrieval for AI agents. One SQLite file, no server, no API key.
FactMem
Score 78A local memory engine. You own the SQLite file. Not Mem0's hosted OpenMemory MCP.
SQLite Local AI Memory
Score 71100% In-Process Bare-Metal Hybrid Memory Engine & MCP Server for LLMs (SQLite + Int8 Vectors).
Five MCP servers worth your time, every week
The Plug Weekly is a free email picked from these scores. About a 3-minute read. Unsubscribe any time.