Shared memory MCP server for AI coding agents (Claude Code, Cursor). Free & local by default.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add threadctx --env THREADCTX_API_KEY='<THREADCTX_API_KEY>' -- npx -y threadctx-mcp~/.codex/config.toml
[mcp_servers.threadctx]
command = "npx"
args = ["-y", "threadctx-mcp"]
env = { THREADCTX_API_KEY = "<THREADCTX_API_KEY>" }
# codex mcp add threadctx --env THREADCTX_API_KEY='<THREADCTX_API_KEY>' -- npx -y threadctx-mcp~/.cursor/mcp.json
{
"mcpServers": {
"threadctx": {
"command": "npx",
"args": [
"-y",
"threadctx-mcp"
],
"env": {
"THREADCTX_API_KEY": "<THREADCTX_API_KEY>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"threadctx": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"threadctx-mcp"
],
"env": {
"THREADCTX_API_KEY": "<THREADCTX_API_KEY>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"threadctx": {
"command": "npx",
"args": [
"-y",
"threadctx-mcp"
],
"env": {
"THREADCTX_API_KEY": "<THREADCTX_API_KEY>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"threadctx": {
"command": "npx",
"args": [
"-y",
"threadctx-mcp"
],
"env": {
"THREADCTX_API_KEY": "<THREADCTX_API_KEY>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"threadctx": {
"command": "npx",
"args": [
"-y",
"threadctx-mcp"
],
"env": {
"THREADCTX_API_KEY": "<THREADCTX_API_KEY>"
}
}
}
}Environment variables
THREADCTX_API_KEY
Replace the placeholders with your own credentials and keep them out of version control.