MCP
Two tools. That's the whole integration.
The CognoDB MCP server exposes exactly two capabilities to an agent: read the schema, run a query. It turns out that's all a capable model needs — it discovers what's in the graph and composes its own questions.
schema
query
Read-only mode
{
"mcpServers": {
"cognodb": {
"command": "npx",
"args": ["-y", "@cognodb/mcp"],
"env": {
"COGNODB_URI": "bolt+ssc://db-7f3a2c1e.databases.cognodb.cloud",
"COGNODB_PASSWORD": "${DB_PASSWORD}"
}
}
}
}Works in Claude Desktop, Claude Code, Cursor, Windsurf, Cline, Zed, JetBrains, Warp and the Gemini CLI — anything that speaks MCP.
Clients
Wherever your agents already run.
The pattern
A graph per agent, reachable over MCP.
Instances start in milliseconds and bill by the second, so each agent or session can own a private, persistent, queryable memory — and you can open it in the console and see exactly what it learned.
Start now
~98.7%
token efficiency at 2,000 entities — see the footnotes above
Point an agent at a graph today.
Free tier, one config block, no integration code.
First-graph path
LiveCreate a free instance
No card. Ready in about a minute.
Connect your driver
bolt+ssc:// URI into the driver you already use.
Write two MERGEs
That's the entire shape of agent memory.
Point an agent at it
One MCP config block. No integration code.