Skip to main content

OpenClaw Memory: How to Give Claude Persistent Memory Across Sessions

· 5 min read
Felo AI
Felo AI
Operations

OpenClaw memory (MemClaw) gives Claude Code persistent project memory across sessions. Install via the Claude Code plugin marketplace — no JSON config required.

Every developer who uses Claude Code hits the same wall eventually.

You open a new session. Claude has no idea who you are, what you're building, or what you decided last week. You spend the first 15 minutes re-explaining everything before you can get any real work done.

OpenClaw memory solves this. MemClaw is the memory skill that makes Claude actually remember your projects — across sessions, across days, across everything.

Get MemClaw free → memclaw.me

image

What Is OpenClaw Memory?​

MemClaw is a skill for AI coding agents — "the AI agent's external brain." It gives Claude persistent, project-scoped workspaces that survive across sessions.

Each workspace stores everything Claude needs to know about a project:

  • Tasks — what's in progress, what's done
  • Artifacts — research reports, docs, URLs, files
  • A living README — background, preferences, current progress
  • Full session history — searchable across time

The "OpenClaw" name comes from the skill ecosystem built around Claude Code. MemClaw is the top-rated memory skill in that ecosystem.

Why Claude Forgets (And Why It Matters)​

Claude Code is stateless by design. Each session is a clean slate. This is fine for isolated tasks, but creates real friction for ongoing work:

  • You spend 10+ minutes re-explaining your project at the start of every session
  • Claude suggests approaches you already rejected
  • Architectural decisions made last week are invisible to Claude today
  • Switching between projects means constant re-orientation

Without memory, you're not getting an AI assistant. You're getting an AI that needs to be briefed from scratch every single time.

How to Install MemClaw​

There are four ways to install, depending on your setup.

Option 1: Claude Code Plugin Marketplace ​

/plugin marketplace add Felo-Inc/memclaw
/plugin install memclaw@memclaw

Option 2: ClawHub​

clawhub install memclaw

Option 3: OpenClaw​

bash <(curl -s https://raw.githubusercontent.com/Felo-Inc/memclaw/main/scripts/openclaw-install.sh)

Option 4: Manual​

git clone https://github.com/Felo-Inc/memclaw.git
cp -r memclaw/memclaw ~/.claude/skills/

MemClaw installs to ~/.claude/skills/ for Claude Code. It also works with Gemini CLI (~/.gemini/skills/) and Codex (~/.codex/skills/).

Setting Up Your API Key​

MemClaw uses your Felo API key. Get one at felo.ai → Settings → API Keys.

Then set it as an environment variable:

# macOS / Linux
export FELO_API_KEY="your-api-key-here"

# Windows (PowerShell)
$env:FELO_API_KEY="your-api-key-here"

Or store it permanently in ~/.memclaw/env. No JSON config files. No MCP setup required.

image

Using MemClaw: Natural Language Only​

Once installed, you interact with MemClaw entirely through natural language — no commands to memorize, no config to edit.

Create a workspace:

"Create a workspace called MyProject"

Load a workspace:

"Load the MyProject workspace"

Query your workspace:

"What's in my workspace?"
"What did we decide about authentication last week?"

Save to workspace:

"Save that report to the workspace"
"Add this decision to the workspace: we're using optimistic locking for inventory updates"

Claude handles the rest — creating tasks, saving artifacts, updating the README — silently in the background as you work.

What Gets Stored​

MemClaw organizes workspace memory into three layers:

Tasks — Claude auto-tracks substantive actions. You can also ask it to create or update tasks explicitly.

Artifacts — Research reports, documentation, URLs, code snippets, anything worth keeping. Stored and searchable.

README memory — The living document of your project: background context, your preferences, current progress, key decisions. This is what Claude reads at the start of each session to get up to speed instantly.

Multi-Project Setup​

Each project gets its own workspace. The workspace registry lives at ~/.memclaw/workspaces.json and maps workspace names to IDs.

Switching projects is just:

"Load the Acme Client workspace"

Claude immediately has the right context. No manual loading, no risk of cross-project contamination.

OpenClaw Memory vs. CLAUDE.md​

Claude Code has a built-in mechanism for persistent context: the CLAUDE.md file. It works, but it has limits:

  • Static — you have to manually update it every time something changes
  • No search — you can't query it, only read it top to bottom
  • No history — decisions and rationale aren't tracked over time
  • No team sharing — it's a local file

MemClaw is dynamic. It accumulates context automatically, supports conversational queries, tracks history with timestamps, and supports shared workspaces for teams.

For a solo developer on one simple project, CLAUDE.md is fine. For anyone managing multiple projects or working on anything long-running, MemClaw is the better tool.

Getting Started​

  1. Install: /plugin marketplace add Felo-Inc/memclaw then /plugin install memclaw@memclaw
  2. Get your API key at felo.ai/settings/api-keys
  3. Set export FELO_API_KEY="your-api-key-here"
  4. Tell Claude: "Create a workspace called [your project name]"
  5. Add context: "Save our tech stack and architecture decisions to the workspace"

From the next session on, Claude loads your workspace automatically. You stop re-explaining and start building.

Start with MemClaw free → memclaw.me