Open source. Self-host or use Lobu Cloud.

Give every agent the context of your company.

Bring ChatGPT, Claude, Codex, OpenClaw, or any agent. Give each one durable company context, scoped access, human approvals, and a complete audit trail.

Try Lobu

app.lobu.ai/acme/activity

AAcmeRecent

Search…Agent scoutApproval: PendingFilters Sort: Newest first4,545 events

3 days agoCreate Issueon GitHub (App install)Requested by renewal-risk-monitorTitle *Acme renewal at risk, no exec championLabels Add items...ConfirmReject

append-only1 awaiting approval

Open Lobu and start building →

#revenue

Tony@slackbot is Acme at risk?

slackbot APPRenewal moved to Jun 30 and a P1 reopened.

Add to Slack →

app.tssdk

import { searchMemory } from "@lobu/client"

const { data } = await searchMemory({

path: { orgSlug: "acme" },

body: { query: "renewals at risk" },

})

Client SDK docs →

claude code

> Which customers are at risk this quarter?

lobu · mcp · reading memory

Acme Corp · 3 signals · 1 renewal

▸▸ auto Opus 5 · 1M context

Install to your Claude →

Works with

ChatGPT Claude Codex OpenClaw Any MCP client

Connects to

DiscordGitHubGmailGoogle CalendarGoogle ChatHacker NewsLinearMicrosoft TeamsPostgreSQLProduct HuntRedditRSS / AtomSlackTelegramWhatsApp CloudX (Twitter)YouTube+3 more

[ 01/THE MULTIPLAYER OPERATING LAYER ]

The context agents need. The control people keep.

Sources stay inspectable, access stays scoped, and consequential actions stay under policy.

One current company state

Connect systems once. Resolve their events into current, source linked company context.

Identity follows the user

Each request inherits the right user, team, device, credential, and task scope.

Govern action, not intent

Agents prepare the work. Policy decides when a person must approve the exact action.

Every outcome becomes context

Approvals, corrections, and completed actions become durable context for whoever works next.

[ 02/BEHAVIORS ]

Turn new events into repeatable work.

A Behavior is a versioned job that wakes on an event or schedule, reads only what is new, and saves or proposes the next outcome.

Renewal risk

Keep every agent current as renewal signals arrive.

salesforce.opportunity.updated + slack.message

Wakes on a relevant change

A renewal date moves or a customer asks for an executive sponsor.

account_context

Reads only the new window

The Behavior combines unprocessed events with a named, read-only SQL view of the account.

customer_risk + approval.requested

Saves a governed outcome

It updates shared account risk and proposes the follow-up task for your approval.

[ 03/TWO WAYS IN ]

Bring any agent. Add shared context.

Use the agents your team already trusts. Managed agents are optional.

Start here

Bring your agents

Connect ChatGPT, Claude, Codex, OpenClaw, or any MCP client to the same company context, scoped to each user.

Connect an agent →

Optional runtime

Build a managed agent

Run repeatable work from live events or schedules under the same approval policy.

Build a managed agent →

[ 04/UNDER THE HOOD ]

One control plane. Any agent.

Share context, permissions, approvals, and connected actions with any agent. Add Lobu's runtime only when you need it.

Halftone figure: inference orb with orbiting nodes

Halftone figure: sealed container with a keyhole

Halftone figure: clipboard checklist with checkmarks

Halftone figure: freight container on a crane hook

Halftone figure: oscilloscope with waveform

01·CONNECTORS

Plug into everything.

20+ connectors (Slack, GitHub, Gmail, Linear, Postgres and more) stream events in, and actions flow back out under scoped identities. Anything with an API fits the connector SDK.

Connect your first agent to company context.

Connect any agent to shared context and keep consequential actions under approval.

Status:// connecting agent…

For humans For agents