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

Gomaa

Gomaa is a Python-based persistent-memory system for AI agents. It organizes memory in human-readable Obsidian Markdown vaults, offers hybrid search and knowledge-graph retrieval, and exposes memory functions through an MCP server. Its documented storage options are local SQLite WAL or PostgreSQL with pgvector.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Gomaa — 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 Gomaa when an agent workflow needs persistent, inspectable memory and you want to choose between a local SQLite setup and a PostgreSQL-backed fleet architecture.

Who it is for

  • Developers adding persistent recall to AI-agent workflows.
  • Obsidian users who want agent memory represented as editable Markdown notes.
  • Teams running multiple agents that need shared memory alongside private stores.

Documented uses

  • Store agent knowledge in a persistent, human-readable vault.
  • Retrieve relevant stored information through semantic, keyword and graph-based search.
  • Separate memory by domain or topic and preserve selected pinned items.
  • Expose memory capabilities to an agent through an MCP server.
  • Configure a shared memory database for a multi-agent fleet.
Before you choose

Requirements & limits

The repository is public and specifies the Apache-2.0 license. It documents Python package and Docker Compose setup paths.

  • The standalone path uses SQLite WAL; PostgreSQL with pgvector is documented for larger search workloads and fleet sharing.
  • MCP is the documented agent interface, with example configuration blocks for Claude Desktop and Cursor.
  • Obsidian-compatible Markdown notes support human review and curation of stored memories.
  • The README lists optional production extras for pgvector, fastembed, server and Google Drive synchronization.
  • Python 3.9 or later is specified.

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 Gomaa and initialize a local memory vault

Set up the lightweight SQLite-based local mode and create a Gomaa vault.

Before you start

  • Python 3.9 or later
  • A terminal and pip

Steps

  1. Install the package with the command below.

    Documented command
    pip install gomaa

  2. Initialize the local vault with gomaa init.

  3. Optionally launch the knowledge-graph dashboard with gomaa dashboard.

Expected outcome

A local Gomaa vault is initialized; the documented dashboard command can launch its knowledge-graph view.

Guide limitations

  • This guide covers local setup only, not connecting an agent or storing and retrieving memories.
  • The repository describes SQLite WAL as the standalone mode; no external database setup is included.
Documentation 1 ↗
Source-based practical guide

Prepare Gomaa's PostgreSQL fleet stack with Docker Compose

Start the documented PostgreSQL-based Gomaa stack for a fleet deployment.

Before you start

  • Git
  • Docker Compose

Steps

  1. Clone the Gomaa repository with the command below.

    Documented command
    git clone https://github.com/M4F-S/gomaa.git

  2. Change into the repository directory with cd gomaa.

  3. Run the command below.

    Documented command
    docker compose up -d

Expected outcome

The README's Docker Compose deployment command starts the documented PostgreSQL, pgvector and Gomaa MCP server stack.

Guide limitations

  • The README presents this path for multi-agent fleets and production servers; this guide does not cover agent configuration or validate a running deployment.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
M4F-S/gomaa · Official homepage title · retained in its source language
How we review sourcesSuggest a correction