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

Kyros

Kyros is an open-source persistent-memory system for AI agents. It organizes records as episodic events, semantic facts and procedural skills, and provides memory decay, cryptographic integrity features and contradiction resolution through a model-agnostic REST API and Python and TypeScript SDKs.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Kyros — 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 Kyros when an agent needs a separate persistent memory layer for events, facts or procedures, and you want to connect through its REST API or an SDK. It is particularly relevant when memory decay, integrity proofs or conflict handling matter to the design.

Who it is for

  • Developers building AI agents that need to retain and retrieve information across interactions.
  • Teams that want to distinguish remembered events, facts and procedures in an agent memory layer.
  • Developers using Python or TypeScript who want SDK access, or teams that prefer a documented Docker Compose self-hosting path.

Documented uses

  • Store an agent’s experiences, such as user conversations or observations.
  • Record and query semantic facts such as a user preference or subscription plan.
  • Represent task skills and workflows as procedural memories.
  • Retrieve relevant stored memories for an agent query.
  • Apply memory aging and inspect integrity or conflicting beliefs through the system’s documented features.
Before you choose

Requirements & limits

The repository is public and licensed under Apache-2.0. The README documents self-hosting and a managed sandbox client option; it does not state pricing or account requirements.

  • Self-hosting requires Docker and Docker Compose; SDK use requires Python 3.11+ or Node.js 18+.
  • The documented SDK choices are Python and TypeScript, with a model-agnostic REST API also available.
  • The README describes SHA-256 and Merkle-tree integrity features, Ebbinghaus-style decay and belief propagation.
  • LangChain and LlamaIndex are listed as integration ecosystems; the README also contains a LangChain example.

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

Set up a self-hosted Kyros server

Start the documented Kyros server locally with Docker Compose.

Before you start

  • Docker and Docker Compose
  • A shell with Git available to clone the repository

Steps

  1. Clone https://github.com/Kyros-494/kyros-ai.

  2. Change into the kyros-ai repository directory.

  3. Run docker compose up -d.

  4. Use the documented local server address, http://localhost:8000, as the client base URL.

Expected outcome

The documented setup starts the server locally at http://localhost:8000.

Guide limitations

  • This covers server setup only; it does not configure an agent or demonstrate a memory query.
  • The README does not detail additional production deployment configuration in the cited setup steps.
Documentation 1 ↗
Source-based practical guide

Install a Kyros SDK

Add the documented Python or TypeScript Kyros client package to a development environment.

Before you start

  • Python 3.11+ for the Python SDK, or Node.js 18+ for the TypeScript SDK

Steps

  1. For Python, run pip install kyros-sdk.

  2. For TypeScript, run npm install @kyros.494/sdk.

Expected outcome

The selected Kyros SDK package is installed for use in the project.

Guide limitations

  • Installation alone does not configure a server connection or demonstrate storing and querying memories.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
Kyros-494/kyros-ai · Official homepage title · retained in its source language
How we review sourcesSuggest a correction