Add the Cognee MCP server to VS Code
Config last verified Nov 10, 2026
The exact config to run Cognee in VS Code — paste it in, restart, and the tools load.
Prerequisites
- VS Code installed.
- LLM_API_KEY — API key for the LLM provider Cognee uses to extract the graph and answer queries (OpenAI by default).
Setup
1. Open .vscode/mcp.json
2. Add this configuration
Add to .vscode/mcp.json
.vscode/mcp.json
json
{
"servers": {
"cognee": {
"type": "stdio",
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"TRANSPORT_MODE=stdio",
"-e",
"LLM_API_KEY",
"cognee/cognee-mcp:main"
],
"env": {
"LLM_API_KEY": "<LLM_API_KEY>"
}
}
}
}Heads up
- VS Code uses the `servers` key (not `mcpServers`) and requires `type`.
3. Restart VS Code and confirm the Cognee tools load.
Gotchas
- VS Code uses the `servers` key (not `mcpServers`) and requires `type`.
VS Code is the exception to every other client: its top-level key is "servers", not "mcpServers", and each entry needs an explicit "type" (for example "http" or "stdio"). Secrets are not inlined; instead you declare an "inputs" array and reference each value as ${input:id}, so VS Code prompts for it once and never writes it to disk. Remote servers connect natively with OAuth.