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

Metronix Memory

Metronix Memory is self-hosted AI-agent memory infrastructure exposed as an MCP memory server, combining durable workspace- and agent-scoped facts with hybrid retrieval over files and SaaS knowledge.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Metronix Memory — 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 it when an agent needs self-hosted, persistent memory and hybrid retrieval beyond a vector database or conversation transcript alone.

Who it is for

  • Developers adding persistent memory to AI agents.
  • Teams building assistants that need to retrieve knowledge from files and SaaS sources.
  • Builders maintaining project knowledge that should stay useful as projects change.

Documented uses

  • Store durable facts, preferences and pinned context for individual agents and workspaces.
  • Retrieve relevant knowledge using dense, sparse and graph context.
  • Provide agents with persistent context across sessions.
Before you choose

Requirements & limits

The repository is public and lists an Apache-2.0 license. The documented quick start requires Docker, at least 6 GB RAM (8 GB recommended) and about 15 GB free disk.

  • The deployment is self-hosted and uses a Docker Compose stack.
  • Retrieval combines dense search, SPLADE sparse search and Neo4j graph context.
  • MCP is the primary integration surface; REST and OpenAI-compatible APIs are also described.
  • Bundled local models are documented, with external answer generation optional.
  • Reported benchmark results have caveats, including directional single runs and external results that are not reproduced in the repository.

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

Install the agent-memory stack

Install the default Metronix agent-memory mode and check the documented local health endpoint.

Before you start

  • Docker
  • At least 6 GB RAM; 8 GB recommended
  • About 15 GB free disk

Steps

  1. Clone the repository from https://github.com/mtrnix/metronix-memory.git.

  2. Change into the metronix-memory directory.

  3. Run ./install.sh to install the default agent-memory mode.

  4. Check the documented health endpoint at http://localhost:8000/health.

Expected outcome

The documented setup flow installs agent-memory mode and provides a local health-check endpoint.

Guide limitations

  • The quick-start instructions point to separate installation documentation for environment configuration, ports and troubleshooting.
  • Connecting an agent is a separate next step; this recipe does not configure an agent.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
mtrnix/metronix-memory · Official homepage title · retained in its source language
How we review sourcesSuggest a correction