Local memory engine for AI agents with knowledge graphs, forgetting, and semantic recall
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add alaya --env ALAYA_LLM_API_KEY='<ALAYA_LLM_API_KEY>' -- npx -y alaya-mcp~/.codex/config.toml
[mcp_servers.alaya]
command = "npx"
args = ["-y", "alaya-mcp"]
env = { ALAYA_LLM_API_KEY = "<ALAYA_LLM_API_KEY>" }
# codex mcp add alaya --env ALAYA_LLM_API_KEY='<ALAYA_LLM_API_KEY>' -- npx -y alaya-mcp~/.cursor/mcp.json
{
"mcpServers": {
"alaya": {
"command": "npx",
"args": [
"-y",
"alaya-mcp"
],
"env": {
"ALAYA_LLM_API_KEY": "<ALAYA_LLM_API_KEY>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"alaya": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"alaya-mcp"
],
"env": {
"ALAYA_LLM_API_KEY": "<ALAYA_LLM_API_KEY>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"alaya": {
"command": "npx",
"args": [
"-y",
"alaya-mcp"
],
"env": {
"ALAYA_LLM_API_KEY": "<ALAYA_LLM_API_KEY>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"alaya": {
"command": "npx",
"args": [
"-y",
"alaya-mcp"
],
"env": {
"ALAYA_LLM_API_KEY": "<ALAYA_LLM_API_KEY>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"alaya": {
"command": "npx",
"args": [
"-y",
"alaya-mcp"
],
"env": {
"ALAYA_LLM_API_KEY": "<ALAYA_LLM_API_KEY>"
}
}
}
}Environment variables
ALAYA_LLM_API_KEY
Replace the placeholders with your own credentials and keep them out of version control.