Nexo runs on your own machine — the client starts it, so there is no endpoint to ping. 892 installs a week from npm.
Cognitive memory for AI agents. Atkinson-Shiffrin, RAG, knowledge graph. 100+ tools.
Nothing serious here today
Today is the operative word: we check Nexo every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add nexo -- npx -y nexo-brain{
"mcpServers": {
"nexo": {
"args": [
"-y",
"nexo-brain"
],
"command": "npx"
}
}
}[mcp_servers.nexo]
command = "npx"
args = ["-y", "nexo-brain"]{
"mcpServers": {
"nexo": {
"args": [
"-y",
"nexo-brain"
],
"command": "npx"
}
}
}{
"mcpServers": {
"nexo": {
"args": [
"-y",
"nexo-brain"
],
"command": "npx"
}
}
}Alternatives to Nexo
same job, measured the same way97-domain knowledge graphs for AI agents — 4× F1 of RAG, 11× fewer tokens. MCP-native.
Persistent knowledge memory for AI agents. Hybrid search, code graph, pgvector.
Two-layer memory MCP server for AI agents with 37 tools, RAG, graphs, wiki, auth
Governed graph-native agent memory: knowledge extraction, fusion, hybrid RAG, scoped access tokens.
Persistent memory and knowledge graph for AI assistants — keyword + vector + graph search.
Offline codebase knowledge graph: 91% token reduction vs naive RAG, plus cross-session agent memory.
Rag Knowledge Graph MCP Server by MEOK AI Labs
Salesforce AgentForce knowledge graph — 40 nodes, Einstein Trust Layer, 11x fewer tokens than RAG.
Nexo — questions
Answers built from our own checks of this server.