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

Iai Personal Memory Engine

Iai Personal Memory Engine is a local MCP server that records AI-agent conversations verbatim, organizes them into episodic, semantic and procedural memory, and provides relevant past context to compatible assistants across turns.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Iai Personal Memory Engine — 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 a compatible agent workflow needs persistent, locally stored conversation context rather than relying on a fresh prompt each session.

Who it is for

  • Users of MCP-compatible AI coding assistants who want context to carry across sessions.
  • Developers configuring local memory and conversation capture for supported agent hosts.

Documented uses

  • Retain exact details from earlier agent conversations for later recall.
  • Give an assistant a relevant memory prefix at the start of a session and context updates on later turns.
  • Organize related past conversations into summaries and learned behavioural parameters.
Before you choose

Requirements & limits

The project is MIT-licensed. Python 3.11 and 3.12 are supported; the README lists macOS and Linux as supported and Windows as beta.

  • Automatic capture and context injection depend on hooks provided by the agent host.
  • The recall path is described as local and LLM-free; an optional idle-time REM consolidation step can invoke Claude through the user's existing subscription.
  • The macOS/Linux source installer requires Git, Python 3.11 or 3.12, Node.js 18 or later, and Rust.
  • The README describes the local store as combining encrypted records, a vector index and a graph; legacy SQLite stores can be migrated to the native engine.

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 package and configure capture for Codex

Set up the package, background daemon and Codex capture hooks using the documented other-host flow.

Before you start

  • Python 3.11 or 3.12
  • An MCP-over-stdio host; automatic capture depends on host hooks

Steps

  1. Install the package with the command below.

    Documented command
    python3.12 -m pip install -U iai-pme

  2. Initialize the local encryption setup with iai-mcp crypto init.

  3. Install the background service with iai-mcp daemon install.

  4. Install capture hooks for Codex with iai-mcp capture-hooks install --target codex.

Expected outcome

The documented package, daemon and Codex capture-hook setup is in place; the host's available hooks determine automatic capture and context injection.

Guide limitations

  • The steps document setup, not a verification of successful installation.
  • MCP tools can work with MCP-over-stdio clients, but automatic capture and context injection depend on host hooks.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
CodeAbra/iai-personal-memory-engine · Official homepage title · retained in its source language
How we review sourcesSuggest a correction