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

Cavemem

Cavemem is a TypeScript project that captures observations from coding-assistant sessions, compresses them and stores them in local SQLite for retrieval across assistant integrations. It provides MCP search, timeline and observation-retrieval tools, with hybrid keyword and local-vector search.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Cavemem — 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 Cavemem when you want supported coding assistants to retain and query local session history across sessions, especially when more than one assistant is involved. Check the capability matrix first: some integrations query existing memories but cannot capture their own sessions.

Who it is for

  • Developers who use compatible coding assistants and want session observations retained across sessions.
  • Teams or individuals evaluating a local, shared memory layer for multiple coding-assistant integrations.

Documented uses

  • Capture observations from supported coding-assistant sessions.
  • Search, review timelines and retrieve stored observations through MCP.
  • Browse saved sessions in a read-only local web viewer.
  • Exclude specified directories or strip tagged private text at the write boundary.
Before you choose

Requirements & limits

The README documents installation as a global npm package. The repository describes Cavemem as frozen and no longer in active development; it says the compressed-memory core now ships inside Caveman.

  • Capture and query support differ by assistant; Cursor, Gemini CLI, Antigravity and IBM Bob are query-only.
  • Codex and GitHub Copilot lack a SessionEnd event, while Augment Code lacks UserPromptSubmit.
  • The default design stores memory locally and makes no network calls; remote embedding providers are an optional configuration.
  • The README describes compression as deterministic and says code and paths are preserved byte-for-byte.
  • Windows Claude Code hook capture may depend on Git for Windows' sh being resolvable.

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 Cavemem and wire an assistant

Install Cavemem and configure a listed coding assistant integration.

Before you start

  • Node.js/npm for the global package installation.
  • A coding assistant supported by Cavemem; capture and query capabilities vary by IDE.
  • On Windows with Claude Code hooks, Git for Windows' sh may need to be on the user Path.

Steps

  1. Install the package globally with the command below.

    Documented command
    npm install -g cavemem

  2. Run cavemem install for the default Claude Code integration, or use cavemem install --ide with a listed IDE identifier.

  3. Run cavemem status to inspect integration wiring and embedding backfill status.

Expected outcome

The selected assistant is wired according to its listed capture and query capabilities; the status command reports the configuration.

Guide limitations

  • Cursor, Gemini CLI, Antigravity and IBM Bob are query-only and do not capture their own sessions.
  • Codex and Copilot do not provide a SessionEnd event; Augment does not provide UserPromptSubmit.
  • On Windows, Claude Code hooks can fail to resolve sh if Git for Windows' Git/bin is not on the user Path.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Vector search

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
CavememCapture observations from supported coding-assistant sessions.The README documents installation as a global npm package. The repository describes Cavemem as frozen and no longer in active development; it says the compressed-memory core now ships inside Caveman.
Engram MemoryPersist agent memories across sessions.Linux, macOS, Self-hosted, WindowsThe repository README identifies a BSL 1.1 license. It describes the Community edition as free for internal use, research and self-hosted deployments; commercial or SaaS use requires a license, and each version converts to Apache 2.0 after four years. The self-hosted setup requires Docker, 4 GB or more of RAM and 10 GB or more of storage. Python 3.10 or later is needed only to run the stdio MCP server or CLI tools di
YourmemoryRetain durable facts and preferences across agent sessions.CLI, Self-hosted, ServerThe Python package setup requires obtaining a token through the project website by entering an email and verifying a six-digit code. The README also describes standalone binaries for macOS, Linux and Windows.
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
JuliusBrussee/cavemem · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

Caveman — The README identifies Caveman as the actively developed home of Cavemem’s compressed-memory core.

Explore project →