create_personaLABSCO SUMMARY
The memory store is a separate service, so it outlives any one client.
You run the vector database first and hand the MCP server a URL and an API key, which means several clients can share the same persona memory rather than each keeping its own file. Memories carry an importance score, and the cleanup tool uses it — old and unimportant entries go, so the store does not grow without limit.
What it is
Two parts: a vector database server for storage and similarity search, and an MCP server that exposes persona and memory management on top of it.
What you get
- Personas: create_persona, list_personas, get_persona, update_persona, delete_persona
- Memory: add_memory with a type and an importance score, and search_persona_memories by semantic similarity
- Conversations: add_conversation and get_conversation_history for complete tracking
- Housekeeping: cleanup_persona_memories for old or low-importance entries, and get_persona_stats
- Diagnostics: get_system_health and test_connection
- Cosine, Euclidean and dot-product similarity on the database side, with SQLite metadata persistence
- API key authentication with role-based permissions and multi-tier rate limiting
Requirements
Node.js 18.0.0 or higher and 2GB of available RAM. Set up the database server first — npm run setup:database and npm run generate:api-key in zero-vector/server — then configure the MCP server with ZERO_VECTOR_BASE_URL and ZERO_VECTOR_API_KEY. The setup puts an OpenAI API key in the database server's .env for embeddings; the architecture diagram also shows local transformers as a provider.
Setup effort
Build from source — git clone https://github.com/MushroomFleet/zero-vector-MCP.git
USE CASES
01
Giving an assistant memory that persists between sessions
Create a persona, then add the facts worth keeping with an importance score.
02
Recalling only what is relevant to the current question
Search that persona's memories semantically instead of replaying the whole history.
03
Keeping the store from growing forever
Clean up by age and importance, and check what is left.
04
Checking the connection before blaming the model
Test the link to the vector server and read its health.
13 tools
Listed in project docs. Your client may show a different list.
list_personasget_personaupdate_personadelete_personaadd_memorysearch_persona_memoriesadd_conversationget_conversation_historycleanup_persona_memoriesget_system_healthget_persona_statstest_connectionBEFORE YOU INSTALL
What to know first
Tools you get
13 tools named in the project's documentation.
Requires
- Node.js 18.0.0 or later
- ZERO_VECTOR_API_KEY — the key npm run generate:api-key prints in zero-vector/server
- OPENAI_API_KEY in zero-vector/server/.env — the setup asks for it for embeddings
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
Clone the repository
TERMINAL
git clone https://github.com/MushroomFleet/zero-vector-MCP.git
cd zero-vector-MCP1. Set up the Zero-Vector server
TERMINAL
cd zero-vector/server
npm install
npm run setup:database
npm run generate:api-key — keep the key it prints; it goes in ZERO_VECTOR_API_KEY belowcp env.example .env — then add your OpenAI API key to this .env
TERMINAL
npm start — the vector database server answers on http://localhost:3000; leave it running2. Set up the MCP server (in a new terminal)
TERMINAL
cd MCP
npm installAdd the Cline configuration under MCP Clients, with the full path of MCP/src/index.js in your clone and the key from npm run generate:api-key
MCP Clients
Cline
CLINE · JSON
{
"mcpServers": {
"zero-vector": {
"command": "node",
"args": [
"<YOUR_PATH_TO_ZERO_VECTOR_MCP>/MCP/src/index.js"
],
"env": {
"ZERO_VECTOR_BASE_URL": "http://localhost:3000",
"ZERO_VECTOR_API_KEY": "<YOUR_ZERO_VECTOR_API_KEY>"
}
}
}
}<YOUR_PATH_TO_ZERO_VECTOR_MCP> is the full path of your clone, and <YOUR_ZERO_VECTOR_API_KEY> is the key npm run generate:api-key printed. The vector database server (npm start in zero-vector/server) has to be running before Cline starts this.
