Start Free — No Card Required — 5,000 observations · 2 devices · Full sync · £0 forever Get started
Free beta is live

The Only AI Memory That Syncs Across Devices, Agents, and Teams

Start a session in OpenClaw on your Mac. Continue in Claude Code on your laptop. Finish in Codex in the terminal. One shared memory layer.

No competitor does this. Supermemory: $20/mo, single device. Mem0: $2/mo, single device. Engrm Free: £0, 2 devices, cross-agent sync.

Quick Setup for OpenClaw

1
Install plugin
openclaw plugins install engrm-openclaw-plugin Click to copy
2
Restart OpenClaw
Quit and reopen OpenClaw (or restart gateway)
3
Connect in chat
/engrm connect Click to copy
Browser opens → Sign in → Return to OpenClaw
4
Verify
/engrm status Click to copy
✓ Startup memory briefs  •  ✓ Auto-save sessions  •  ✓ Cross-device sync
TypeScript · Fair Source (FSL-1.1-ALv2) · OpenClaw native · Offline-first
Cross-device sync (no one else has this)
OpenClaw · Claude · Codex ready
Team memory
Delivery Review
Sentinel
< 50ms
Local search latency
384-dim
Offline semantic vectors
FTS5 + Vec
Hybrid search (RRF)
Cross-device sync: Mac to Cloud to Laptop
OpenClaw (Mac)
Planning, research, daily work
Shared Engrm Memory
Briefs, decisions, lessons, next steps
Claude / Codex (Laptop)
Coding, implementation, review
Live knowledge map observations cluster by type, project, timeline, and author

Free forever for solo devs. Upgrade when you need more.

  • Hit 5,000 observations? → Vibe (£5.99) gives you 25,000
  • Need a 3rd device? → Vibe supports 3 devices
  • Want blocking Sentinel? → Team (£12.99/seat) stops risky code automatically
  • Building with a team? → Team plan adds shared namespace + unlimited observations

Why Engrm?

Better free tier. Better Pro tier. Better team tier.

Feature Engrm Free Supermemory Pro Mem0
Cost £0 $20/mo ~$2/mo + embeddings
Cross-device sync ✅ 2 devices ❌ No ❌ No
Team memory ✅ (Team plan) ❌ No 🟡 Coming
Sentinel (risk detection) ✅ Advisory ❌ No ❌ No
Delivery Review ❌ No ❌ No
Observations 5,000 (~3 months) ~150k tokens Pay-per-embed
Live now

Start free today

OpenClaw, Claude Code, and Codex can now share the same memory layer across devices. Team memory, deeper Sentinel workflows, and paid plans are coming next.

Free beta includes 2 devices, 5,000 observations, cross-agent memory, packs, and sync. No credit card required.

How cross-device memory works

Engrm architecture: Agent layer, Capture & Sync, Storage

One namespace. Every agent. Every machine.

1

Capture sessions

OpenClaw sessions become structured observations: decisions, lessons, next steps, delivery evidence.

2

Sync across devices

That memory follows your namespace, not the machine. Mac → Laptop → Team — one shared layer.

3

Surface when needed

Startup briefs, recent activity, and Sentinel checks bring old context back at the right time.

OpenClaw on Mac
|
Claude Code on Laptop   <- shared namespace ->   Codex in Terminal
|
Engrm shared memory
|
briefs + decisions + lessons + follow-ups + delivery evidence
|
team context + personal continuity

Why people switch to Engrm

Most memory tools stop at recall. Engrm is built for continuity and follow-through.

Cross-device shared memory

What happened on your Mac is available on your laptop. The work doesn't restart cold just because the hardware changed.

Shared team memory

Team decisions, discoveries, and lessons become reusable context instead of disappearing into chat history.

Delivery Review when it matters

When the work is code, Engrm compares what was promised with what actually shipped.

See all features →

Featured packs for beta

Packs are free. They give Engrm a shared vocabulary for delivery drift, risky refactors, security mistakes, and frontend traps.

Delivery Review

Plan vs delivered

Helps Engrm decide whether a session actually shipped what the brief called for, or just looked convincing.

Refactor Guard

Refactors with receipts

Highlights when cleanup starts consuming the session without evidence of the promised outcome.

Security Preflight

Catch obvious risk fast

Encodes practical auth and secret-handling lessons that should never need rediscovering.

Explore the pack library

One memory layer. Every agent.

Engrm works with OpenClaw, Claude Code, Codex, Cursor, and every MCP-compatible agent

OpenClaw
Claude
Codex
Cursor

+ Windsurf · Cline · Zed · VS Code Copilot · Any MCP agent

Works with your agent

One memory layer. Every agent.

OpenClaw
Claude Code
Cursor
Codex CLI
Windsurf
Cline
Zed
VS Code Copilot
Paid beta coming soon

Sentinel

Real-time AI audit for coding agents. Sentinel intercepts your agent's code changes before they're written and blocks violations automatically.

Claude writes auth middleware
  |
  v
PreToolUse hook fires
  |
  v
Engrm retrieves team standards from memory
  |
  v
Audit LLM reviews diff + standards
  |
  +---> PASS: Agent proceeds
  +---> BLOCK: "MD5 detected. Use bcrypt."
Learn More About Sentinel

Simple pricing

Upgrade path: Free to Vibe to Pro to Team

Start free. Upgrade when you hit limits.

Free

£0/mo
OpenClaw-native memory, no card required
  • 5,000 observations
  • 2 devices
  • Cloud sync included
  • Local FTS5 search
  • Secret scrubbing
  • 2 help packs
  • 30-day retention
Get Started

Why upgrade?
→ Hit volume limit or need 3rd device

Pro

£9.99/mo
For power users with deeper continuity
  • 100,000 observations
  • 5 devices
  • Unlimited agents per device
  • Hybrid search (local + cloud)
  • Deeper Delivery Review
  • Longer-lived project briefs
  • All help packs
  • 1-year retention
  • Sentinel advisory mode
  • 200 audits/day
Notify Me

Why upgrade?
→ Want blocking Sentinel or shared memory

Team

£12.99/seat/mo
Shared knowledge + Sentinel (min 3 seats)
  • Unlimited observations
  • Unlimited devices
  • Shared decisions and briefs
  • Team Delivery Review
  • All help packs
  • Unlimited retention
  • Sentinel blocking mode
  • Unlimited audits
  • Shared team namespace
Notify Me