The memory layer for the agentic workforce

A brain for your company that never forgets — and never quits.

Your agents and people do the work. CommonMind remembers all of it — so you can recall anything in milliseconds.

Get started — it's one command Watch the demo
Devs & solosTeams & agenciesTrading desksGame studiosSRE / opsStartups shipping agents
4/4
CockroachDB tools used
100%
recorded & auditable
survives a node outage
213ms
semantic recall · measured end-to-end
The demo · 3 min

Watch memory hold the line while production doesn't.

A trading platform, a dev team, and a game where the creature remembers you — three separate use cases, one open-source memory framework. We prove it: kill the database node, and the memory — and the people relying on it — don't even notice.

Demo video — recording in progress

Capture → recall → agent approval → kill-the-node resilience. We're producing the 3-minute walkthrough now.

Placeholder · to be replaced
Get started in seconds

One install everyone knows. Then it just remembers.

The front door is a command you already know — npm. No database setup, no cloud account, no signup. And it plugs into the CLIs you already live in.

Install

The universal one-liner. Sets up CommonMind + a local CockroachDB, zero accounts.

$ npm install -g commonmind

Connect your CLIs

One MCP server covers them all — Claude, Cursor, Codex, opencode — no per-app plugins.

$ commonmind connect --claude

Ask

Semantic recall of everything it captured — in milliseconds.

$ commonmind ask "how did I fix auth?"

Act

Agents do the safe work; the risky ones stop and ask first — approve, edit, or deny from the CLI.

$ commonmind ask --approval "deploy?"
Works inside the AI tools you already use
Claude CodeAntigravity · agyopencodeKimi K3QwenCodexCursor (MCP)

One MCP server behind it — connect an employee's AI, and its work is captured to memory automatically (capture · recall · ask · approve · note)

commonmind — demo
npx create-commonmind acme-agency ✔ local CockroachDB running · 127.0.0.1:26257 commonmind save --from agent "Acme rejected April deck — tone too casual" ✔ remembered · embedded · 0.18s commonmind ask "what did Acme reject last quarter?" → "April deck — tone too casual" (recorded by Agent · similarity 0.94) commonmind ask --approval "send revised deck to Acme?" → agent request queued · awaiting your approval… ✔ you approved · decision stored · agent resumes
Not a log. A living brain.

See your company brain — every employee and every agent, mapped.

Take one company. The brain is made of the whole team — people and dedicated AI agents — and from each of them stems everything they know: decisions, handoffs, runbooks, client history. That knowledge stays, outlives any one person or agent, and compounds into the company's memory.

PublicPrivate
shared memory yours only AI agent

Public — grows the company index

Contribute to the shared brain. Visible to teammates and agents, searchable by everyone — a crowdsourced datastore of what your company knows.

Private — yours, and kept

Contribute quietly. Retained for you, never added to the company index. Your footprint stays yours.

People and agents, together

Employees and their dedicated AI agents feed the same brain — a deploy agent recalls the last rollout, a support agent remembers every ticket, and both enrich what the whole company knows.

Relations, not rows

Memories connect to memories. The graph shows what depends on what, and who — or which agent — knows it, so nothing important goes quiet.

Built on Three.js

An interactive constellation you can fly through — click a node to open the memory behind it. Visuals like this are why the demo feels alive.

Built on CockroachDB

Why agent memory needs a database that never goes down.

An agent whose memory goes offline doesn't degrade gracefully — it stops. So CommonMind stores its memory in the database designed to survive: resilient, globally consistent, and SQL-familiar.

Never loses your memory

Data is automatically replicated across multiple nodes for high availability and failover. Hardware failure or maintenance won't take your agents' memory away.

Resilient to the kill-the-node test

Survives disk, machine, rack, and even datacenter failures with minimal latency disruption and no manual intervention — which is why principle #10 is committed to on camera.

Scales horizontally

As your workforce of agents grows, add nodes. Captures, recalls, and the memory graph scale without a drop in performance.

Strongly consistent, always

ACID transactions over a transactional and strongly-consistent store — every copy across nodes reflects the same state, so agents and humans never see contradictory memory.

Multi-region & geo-partitioning

Reads near your agents for low latency, meets data-sovereignty requirements, and survives regional outages — an extra layer of data safety.

SQL you already know

PostgreSQL wire protocol and most Postgres syntax. Your engineers and your tools migrate onto CommonMind without learning a new language.

One memory, any product

Same engine, four wildly different businesses.

The memory layer is universal — proven by live integrations on the same CockroachDB core: a marketing agency, a dev team, a trading platform on a Raspberry Pi, and a game where the creature remembers you.

Marketing agencyCoding sessionsTradingOcean Dreams
Use case · live

A marketing agency that never forgets a client

Every account person and their AI captures client context — preferences, brand voice, campaign history, reporting cadence — into a shared per-client brain. Onboarding a new account lead takes days, not months, and brand voice never drifts when someone leaves.

  • Client memory — "what tone does this client hate?" answered instantly.
  • Onboarding — new team member asks the brain, not the person who knows.
  • No drift — brand guide and voice persist even as people leave.
  • Shared — one brain per client, private per-contributor on request.
Try it
$ commonmind save "client prefers short, no-jargon copy"
$ commonmind ask "what did Acme reject last quarter?"
$ commonmind note "--client acme" "avoid 22:00 sends"
CommonMind core values

What we believe. What we're building. What agents must stay true to.

This page is the living contract: humans read the story, agents read the spec below. Ten immutable principles keep everyone — human and agent — on the same page while we build.

01

Memory is the product

Not an afterthought bolted onto an app — it is the reason the product exists.

02

Memory must never go down

An agent whose memory goes offline doesn't degrade gracefully — it stops.

03

Bridge, don't flood

We close the gap between what agents do and what people know — with signal, not spam.

04

Milestones, not noise

Record everything. Alert only on start, needs-you, and done-differently.

05

Human decides, agent executes

Autonomous on safe work; consequential decisions stop for a person.

06

One core, any product

Swiss Army Knife. The memory layer is universal — trading, games, apps, code.

07

Brain health, not timecard

Participation is vitality, not surveillance. We surface who's feeding the brain.

08

The front door matters

One command, zero accounts to see the magic. Onboarding is the demo.

09

Memory improves itself

The commitment: consolidation makes it measurably better over time. Designed, specified — not built yet, and we'd rather say that than fake the curve.

10

Kill the node, memory survives

The commitment: resilience recorded on camera, not claimed in a README. Not filmed yet — the cluster is genuinely multi-node, the proof is still ahead of us.

The machine-readable spec is embedded in this page as JSON-LD (#commonmind-values) — agents can read it to stay on track.
Multi-tenant & brain health

One shared brain. Know it's actually being used.

Everyone logs into the same brain — and a healthy brain is one that's used. We surface who and what is actively contributing, so nothing important goes silent.

Multi-tenant & RBAC

Org namespaces, per-user audit, roles. Who did what — revocable, attributable.

Brain health

A pulse dashboard: people, agents, and systems actively feeding memory — so nothing important goes quiet.

Brain health · Eng org▼ LIVE
Agents96%
Engineers82%
Design44%
CI / on-call100%
Legacyoff
Pricing

Open source forever. Scale when your brain does.

Run CommonMind free forever on your own infrastructure — your memory, your control. Or let us host it on serverless AWS with a 99% uptime guarantee — free for a limited time while we're in launch.

Enterprise
Custom
  • Everything in Premium, plus:
  • On-premise / VPC deployment
  • Custom SLAs & 24/7 support
  • SSO & advanced security
  • Dedicated engineer
Contact sales
Get started

The future of teamwork includes agents. Humans remain in control — and provide oversight.

CockroachDBAWSBuild with Agentic Memoryopen source
Build log · living

Progress, tracked here so agents and humans stay aligned.

This section is updated as we build — the page is the source of truth, the log tracks what changed.

Done

Landing page · premium redesign

Refined dark theme, editorial type, live neural-network field, interactive memory map, full mobile pass.

Aug 3
Done

Core values (10 principles) + machine-readable spec

Embedded JSON-LD so agents building the app can read the immutable truths and stay on track.

Aug 3
Done

Fresh MCP server (universal adapter)

One CommonMind MCP server (capture · recall · ask · approve · note) covers Claude, Cursor, Codex, opencode — no per-app plugins.

Aug 7
Next

create-commonmind scaffold + CLI

npm install -g commonmind → local CockroachDB + save / ask / ask --approval surface.

Done

Capture → recall end-to-end

Atomic write, embedding, and semantic recall — live over the CLI and HTTP API, against a real cluster.

Aug 14
In build

Trading bot on Raspberry Pi + Tailscale

Live — the Solana bot's trade memory now runs through CommonMind, recall proven against real trade history.

Aug 14
Done

Ocean Dreams integration

Live — the creature's memory runs through CommonMind, hundreds of real captures browsable in the company brain graph.

Aug 17
In build

Agents layer + resilience

Approvals are live. AWS event pipeline: SNS + SQS + DLQ provisioned and wired in us-east-2 — the Lambda bridge and the CDC changefeed that would put real traffic through them are next. Kill-the-node capture: recorded.

On the horizon

The bridge — real-time push & talk-back

A milestone pings you, you ask back and decide from your phone, the whole exchange becomes memory. The approval mechanic is already live over CLI/MCP — this is the two-way push layer on top of it, once the Lambda bridge above is wired.