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

Memory Lancedb Pro

Memory Lancedb Pro is a LanceDB-backed OpenClaw plugin that stores preferences, decisions and project context for recall across sessions and agents, using hybrid vector and BM25 retrieval with optional cross-encoder reranking and multi-scope isolation.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Memory Lancedb Pro — 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 OpenClaw agent needs persistent, searchable context between sessions and you can configure the plugin, embedding provider and supported host environment.

Who it is for

  • OpenClaw users who want agents to retain preferences and project decisions across sessions.
  • Operators who need to keep memory separate across agents, users or projects.

Documented uses

  • Capture relevant conversation details for later sessions.
  • Recall preferences or the rationale behind past decisions.
  • Search stored memories with vector and BM25 retrieval, with cross-encoder reranking.
  • Manage agent memory through the plugin's CLI.
Before you choose

Requirements & limits

The documented installation uses the OpenClaw CLI or npm and is self-hosted. The README identifies OpenClaw 2026.3+ and an AVX-capable CPU requirement; native vector search may require AVX2 on some Linux x64 builds.

  • It is an OpenClaw plugin, so the documented setup depends on OpenClaw rather than a standalone user interface.
  • The manual configuration assigns the plugin to OpenClaw's memory slot and specifies embedding settings.
  • The README warns that AVX2 may be needed for native vector search on some Linux x64 builds and documents a fallback setting.
  • The retrieval design combines vector and BM25 search and includes cross-encoder reranking.
  • Memory scopes can separate agent, user or project contexts.

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 and configure the plugin in OpenClaw

Add Memory Lancedb Pro to an OpenClaw installation and configure the documented embedding and recall settings.

Before you start

  • An OpenClaw installation; the README identifies version 2026.3+ for its adapted plugin architecture.
  • A CPU supporting AVX; native vector search may require AVX2 on some Linux x64 builds.
  • An embedding provider and its required API key, if using a hosted provider.

Steps

  1. Install with the OpenClaw CLI using openclaw plugins install memory-lancedb-pro@beta, or install through npm with the command below.

    Documented command
    npm i memory-lancedb-pro@beta

  2. If installed through npm, add the plugin installation directory as an absolute path in plugins.load.paths in openclaw.json.

  3. Set memory-lancedb-pro as the plugins.slots.memory value and add an enabled plugin entry with embedding-provider settings.

  4. Configure options such as autoCapture, autoRecall and smartExtraction as needed; the README's example enables them and uses an environment-variable placeholder for the embedding API key.

Expected outcome

The plugin is installed and configured as OpenClaw's memory plugin, with the example settings enabling automatic capture and recall.

Guide limitations

  • This setup sequence does not itself demonstrate successful runtime operation.
  • If npm is used, the plugin path must be configured as an absolute path.
  • CPU instruction support can affect native vector search; the README documents a setting to disable native cosine search for affected systems.
Documentation 1 ↗
Evidence and transparency

Sources & review

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