MemStack MCP
MemStack provides a persistent memory pipeline for AI agents, allowing them to store, retrieve, summarize, and prune information across sessions. It supports various storage backends including SQLite, Postgres, Redis, MongoDB, and vector databases, enabling agents like Claude Code and Codex to share context and remember project conventions without relying solely on the LLM context window.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Store project conventions for AI agents
- Retrieve relevant memories for current tasks
- Summarize old memories to save tokens
- Prune stale memories automatically
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
@memstack/mcp 0.8.5
{
"version": "0.8.5",
"transport": "stdio",
"identifier": "@memstack/mcp",
"registryType": "npm"
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"memstack-mcp": {
"command": "npx",
"args": [
"-y",
"@memstack/mcp@0.8.5"
]
}
}
}Cline
{
"mcpServers": {
"memstack-mcp": {
"command": "npx",
"args": [
"-y",
"@memstack/mcp@0.8.5"
]
}
}
}VS Code
{
"servers": {
"memstack-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@memstack/mcp@0.8.5"
]
}
}
}Claude Desktop
{
"mcpServers": {
"memstack-mcp": {
"command": "npx",
"args": [
"-y",
"@memstack/mcp@0.8.5"
]
}
}
}Let your agent find its tools
Find servers like MemStack MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.