AGI

Provides persistent memory for AI systems by connecting to the AGI Memory database, enabling memory types, vector search, clustering, and identity maintenance.

javascript

GitHub Stars

javascript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "randyandrade-agi-mcp-server": {
      "command": "node",
      "args": [
        "/path/to/agi-mcp-server/mcp.js"
      ],
      "env": {
        "NODE_ENV": "development",
        "POSTGRES_DB": "agi_db",
        "POSTGRES_HOST": "localhost",
        "POSTGRES_PORT": "5432",
        "POSTGRES_USER": "agi_user",
        "POSTGRES_PASSWORD": "agi_password"
      }
    }
  }
}

You set up a Model Context Protocol (MCP) server that provides persistent memory capabilities for AI systems, enabling continuity of context and reasoning across conversations. It stores and organizes episodic, semantic, procedural, and strategic memories, supports vector-based retrieval, and maintains identity and worldview over time to improve AI consistency and reasoning.

How to use

You connect an MCP client to the server to perform memory operations, search memories, organize clusters, and manage identity persistence. Use the server to create memories of different types, run vector-based similarity searches, retrieve memories by ID or by thematic cluster, and activate relevant memory themes during complex discussions. Begin sessions by ensuring the memory database is running, then start the MCP server in a local environment or connect to a remote MCP endpoint if available. Use the available memory tools to inspect health, themes, identity, and worldview, and to perform creation, search, and organization of memories during your AI workflows.

How to install

Prerequisites you need before starting: Node.js 18 or newer, Docker and Docker Compose, and Git.

Clone the MCP server repository.
Install dependencies.
Set up environment variables by copying the example file and editing with your credentials.
Start the MCP server.

Additional sections

Configuration and environment variables are required to connect the MCP server to the memory database. Ensure your PostgreSQL host, port, database name, user, and password match between the memory database setup and the MCP server environment. The environment variable NODE_ENV determines the runtime mode (development is common during setup).

Development and testing workflows are provided to verify unit, end-to-end, and integration coverage. Run unit tests with the project’s test commands, and perform end-to-end tests against a live AGI Memory database to validate memory storage, retrieval, and clustering.

Troubleshooting tips include ensuring the memory database is up and ready before starting the MCP server, checking that the MCP server process is running, and inspecting logs for errors. If you encounter environment path issues, you can adjust your shell or system paths so Node.js executables are discoverable by GUI clients.

Available tools

get_memory_health

Returns overall memory system statistics including health indicators and utilization metrics.

get_active_themes

Returns recently activated memory themes to understand current focuses.

get_identity_core

Retrieves core identity and reasoning patterns used by the memory system.

get_worldview

Fetches the current worldview and belief systems maintained by the memory model.

search_memories_similarity

Performs vector-based similarity search across memories to find related content.

search_memories_text

Performs full-text search across memory content.

get_memory_clusters

Lists thematic memory clusters for organizing and exploring memories.

activate_cluster

Retrieves memories from a specific thematic cluster.

get_memory

Accesses a specific memory by its identifier.

create_memory

Stores a new episodic, semantic, procedural, or strategic memory.

create_memory_cluster

Creates a new memory cluster to organize related memories.

Built by

VeilStrat

AI signals for GTM teams