Built for Claude Code

Project memory that survives every Claude session

Wolbarg Workspace keeps one reconciled view of your repo — decisions, architecture, tasks, bugs — so Claude Code starts informed and leaves the workspace better for next time.

$ npx @wolbarg/workspace connect claude

Connect Claude Code Install Workspace Cursor & Codex too

Independent product for coding agents. Visual language inspired by Claude Code — not affiliated with Anthropic. Wolbarg SDK

Why Workspace

Claude Code forgets. Your project shouldn't.

Sessions end. Context evaporates. Wolbarg Workspace keeps one always-current briefing so the next Claude run starts where the last one left off.

Claude Code first

One command connects Claude Code to shared project truth via MCP — no hand-edited JSON.

Shared across hosts

Cursor and Codex read the same memory plane. A decision in Claude is known everywhere.

Reconciled, not appended

Updates replace stale facts instead of piling up notes.

Evidence-backed

Decisions stay tied to files, tests, and docs.

Quick path

Init once. Connect Claude. Restart.

1. npm install @wolbarg/workspace

2. npx @wolbarg/workspace init

3. npx @wolbarg/workspace connect claude

Claude Code setup Full install docs

Connectors

Claude Code — plus Cursor & Codex

Featured

Claude Code

Primary host — MCP project memory

npx @wolbarg/workspace connect claude

Setup Claude Code

Host

Cursor

MCP + rules for Cursor

npx @wolbarg/workspace connect cursor

Setup Cursor

Host

Codex

Shared truth for Codex CLI

npx @wolbarg/workspace connect codex

Setup Codex

Building a custom agent?

Use the Wolbarg SDK and AI framework adapters on wolbarg.com.

SDK docs