OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Hexus

Hexus is a Python memory layer for AI agents that stores vector memories in PostgreSQL with pgvector and provides access through a Hermes Agent plugin or a standalone Model Context Protocol server.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Hexus — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Hexus when you want to operate agent memory on PostgreSQL with pgvector and integrate it through Hermes Agent or MCP.

Who it is for

  • Hermes Agent operators who want PostgreSQL-backed memory.
  • Teams connecting MCP-compatible clients such as Claude Desktop or Cursor to an agent memory service.
  • Operators of cooperating agents who need shared or identity-scoped memory reads.

Documented uses

  • Add persistent vector memory to a Hermes Agent deployment.
  • Expose Hexus memory capabilities to an MCP-compatible client.
  • Keep agent memories in a PostgreSQL store shared by a fleet or scoped for reads by agent identity.
Before you choose

Requirements & limits

  • The documented database stack is PostgreSQL 16 with pgvector.
  • A local MiniLM-L6-v2 embedder is shared by the plugin and MCP server.
  • MCP transport can be configured as stdio or HTTP.
  • The default shared read mode allows agents to read across the shared knowledge base; strict mode scopes reads to the calling identity.
  • The Compose setup does not publish the MCP port to the host unless its port mapping is enabled.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Configure Hexus as a Hermes Agent memory plugin

Install Hexus and configure Hermes to use it as its PostgreSQL-backed memory provider.

Before you start

  • A Hermes Agent setup
  • PostgreSQL and a connection string
  • Python package installation with pip

Steps

  1. Install the package with the command below.

    Documented command
    pip install hexus

  2. In Hermes configuration, set plugins.memory.provider to hexus and provide the PostgreSQL DSN under its config.

  3. Set the top-level memory.provider to hexus so Hermes selects it as the memory backend.

Expected outcome

Hermes is configured to use Hexus as its memory provider, backed by the supplied PostgreSQL database.

Guide limitations

  • The repository specifies that both configuration blocks are required.
  • The guide documents configuration, not a verified deployment or runtime result.
Documentation 1 ↗
Source-based practical guide

Start the MCP server and PostgreSQL with Compose

Use the repository's Compose MCP profile to start PostgreSQL with pgvector and the Hexus MCP server.

Before you start

  • Docker Compose
  • The repository's docker/compose.yml
  • A database password set in HEXUS_DB_PASS

Steps

  1. Set HEXUS_DB_PASS to the password used by the Compose setup.

  2. Run the command below.

    Documented command
    docker compose -f docker/compose.yml --profile mcp up

Expected outcome

The Compose MCP profile starts the pgvector database and MCP server on the internal Docker network.

Guide limitations

  • The MCP port is not published to the host by default; the Compose port mapping must be enabled for host access.
  • This recipe covers startup only, not client configuration or a verified connection.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
codenamekt/hexus · Official homepage title · retained in its source language
How we review sourcesSuggest a correction