OpenClaw × OpenClone

Your AI agent, finally knows you

Connect your OpenClone memory vault to OpenClaw. Your agent gains full context: past conversations, screen activity, and work patterns. It acts smarter and faster on your behalf.

clawhub install openclone-memory
Setup guide →

From blind to context-aware

Without memory

Your agent forgets everything after each session. It cannot reference past decisions, knows nothing about your preferences, and has to be re-explained from scratch every time.

you › "Who was in my last standup?"
agent › "I don't have access to your meeting history. Could you paste the notes?"
With OpenClone memory

Your agent searches your vault before acting. It knows your work patterns, remembers past conversations, and makes decisions based on your actual history, not a blank slate.

you › "Who was in my last standup?"
→ searching vault: Audio/2026-02-25.md
→ found: standup 9:00am, Alex · Sarah · Marcus
agent › "Alex, Sarah, and Marcus. Alex mentioned the API delay."

What becomes possible

Real agent interactions, powered by your personal memory

Context-aware scheduling

"Schedule a follow-up with whoever I met on Tuesday." Your agent searches your audio vault, finds the name, then books the slot. No further prompting needed.

you › "follow up with the person from Tuesday's call"
→ audio/2026-02-25.md: meeting with Sarah M.
→ calendar: Thursday 3pm slot is free
agent › "Done. Meeting with Sarah booked for Thu 3pm."

Smart email replies

"Draft a reply to the Hetzner invoice." Your agent checks your screen captures, finds the billing context, and drafts a reply that already knows your plan and payment history.

you › "draft a reply to the Hetzner invoice"
→ screen/2026-02-24.md: invoice #2847, CX22
→ found: plan, payment method, previous emails
agent › "Reply drafted with invoice ref and plan details."

Proactive morning brief

Every morning at 7:30am, OpenClaw queries your vault insights and calendar, then sends a personalized brief to WhatsApp or Telegram, before you wake up.

cron: 07:30 daily
→ GET /api/insights → 3 active projects
→ calendar: review with Alex at 2pm
WhatsApp › "Morning! Review w/ Alex at 2pm. Yesterday you were working on the auth flow."

Memory-first research

"What do I already know about Tailscale?" Get an instant synthesis from your own vault before searching the web. Your knowledge, surfaced first.

you › "what do I already know about Tailscale?"
→ vault search: 3 notes found
→ installed Feb 3, exit node configured Feb 12
agent › "Set up 3 weeks ago. Exit node was your last step."

Three steps to connect

1

Install the skill

Run clawhub install openclone-memory in your terminal. The skill registers itself with your OpenClaw agent automatically.

clawhub install openclone-memory
2

Configure your vault

Open the macOS desktop app → OpenClaw tab. Add your backend URL, API key, and choose which operations to allow.

Desktop app → OpenClaw tab
3

Start asking

Your OpenClaw agent can now search your vault, ask your AI, and retrieve insights from any conversation or workflow.

Your vault stays on your server. The skill only queries your own OpenClone instance, over your local network or VPN. Nothing is sent to third parties.

clawhub install openclone-memory
Setup guide →

Requires OpenClaw and an OpenClone backend