Your goal. Your AI crew.

All hands on
your deck.

Your people and AI agents, working together.
One platform. On your infrastructure.

See your crew in action

Your course. Your crew.

Welcome aboard.
Let’s get to work.

One workspace to coordinate your agents and review their work.

crewship / your workspaceActual app · Demo data

The actual Crewship app showing Operations, Finance and Marketing, and Sales and Shipping crews.

Explore the application

THE AGENT PLATFORM. ON YOUR TERMS.

Your people. Your agents.
Your infrastructure.

Run an agent workspace on your own hardware. Give every crew an isolated Linux environment, keep its knowledge with your team, and choose the AI that powers the work.

Run it on your hardware.

Deploy on a company server or your own machine. Connect the systems your team already uses.

A container for every crew.

Separate Linux environments, with the tools, dependencies, and access each crew needs.

Choose your AI.

Use supported providers and models through CLI adapters, or a local model through Ollama.

ONE SHARED WORKSPACE

People and agents.
Working in the same place.

Conversations, issues, routines, and credentials belong to one platform — alongside the memory and journal that keep the work connected.

Chat & inboxKeep conversations with the work.

Issues & projectsGive requests a place and an owner.

RoutinesSchedule work and respond to events.

CredentialsConfigure the access your crew needs.

Self-hosted workspace. Your choice of inference: cloud providers receive the context you send; local model setups can keep inference on your infrastructure. Availability depends on the adapter and configuration.

INSIDE CREWSHIP

One platform.
See the real workspace.

Actual screenshots from our development instance, with demo data.

Give every crew a purpose.
Sales, finance, and operations crews in the actual Crewship workspace.
Captured from dev1 · Demo workspace · September 2026

WHAT’S UNDER THE DECK

Useful once.
Ready for the next time.

The next run starts with the context you kept, the tools you configured, and a routine your team can run again.

SHARED CONTEXT

Your knowledge. On your infrastructure.

Persistent agent memory and shared crew knowledge stay on your infrastructure. Keep procedures, decisions, and experience with your company.

Use a local model when context must stay on-premises. Cloud models receive the context you send.

REPEATABLE WORK

Good work. On repeat.

Schedule recurring work, trigger it with webhooks, and send notifications when a run needs attention. Inspect each run and keep people in the loop.

YOUR CHOICE

Bring the AI you already use.

Choose a supported CLI, connect your own provider credentials, or use a local model. Your team keeps the choice.

Provider and model availability depends on the CLI adapter you choose.

REAL TOOLS. REAL LINUX ENVIRONMENTS.

Your tools.
Their own workspace.

Give your crew the documents, spreadsheets, and data it needs. Each crew works in its own Linux devcontainer, with tools and access configured for the job.

Familiar work apps alongside developer tools
Tools and dependencies configured per crew
Assigned credentials and controlled access

Explore the infrastructure

Your infrastructureIllustrative setup

LINUX DEVCONTAINER

Operations crew

Debian workspacedebian:bookworm-slim

Alex

Jordan

Google DriveMicrosoft 365

Separate workspace · Shared crew context

LINUX DEVCONTAINER

Data crew

Python on Debianpython:3.12-slim

Morgan

Taylor

PythonPostgreSQL

Separate workspace · Shared crew context

Different environments. One traceable workspace.

Built around the way your team works.

Google Drive

Gmail

Google Sheets

Microsoft 365

GitHub

Python

PostgreSQL

Illustrative tool setups. Connect services through configured CLIs, APIs, or MCP tools, using your own credentials and permissions.

BUILT TO BE INSPECTED

Stay in charge of your crew.

Control the access.

Use workspace roles, agent ownership, and crew permissions to control who can manage the work. Scope the credentials each crew receives.

Keep the trail.

Inspect conversations, run events, and tool activity. Follow a report back to the work that produced it.

Own the environment.

Run Crewship on your infrastructure. Inspect the source and review the current release before deploying. Crewship is in open beta.

Usage visibility depends on the provider and adapter. Subscription activity is not an exact per-task bill.

Explore access and oversight

Who’s aboard

Same ship.
Different crews.

Explore eight engineering workflows as one example of what crews can do. Each crew runs in its own container, picks the CLI best for the job, and lands every action in the journal.

Different crews. One place to follow the work.

See all modules

BEFORE YOU SET SAIL

A few things to know.

What do I need to run Crewship?+

A host with a supported container runtime, such as Docker or Podman. Install Crewship, check the environment with crewship doctor, and configure your first crew. The setup guide covers the supported installation paths.

Can I use local AI instead of a cloud provider?+

A supported Ollama setup can run inference on your own infrastructure. You can also use supported cloud providers through CLI adapters. Available models, authentication, and tool support depend on the adapter and configuration.

Does each agent have its own container?+

The runtime boundary is one Linux container per crew. Agents have their own identities and can have persistent memory; agents in the same crew share its working environment.

Does company knowledge stay on our infrastructure?+

The self-hosted workspace stores memory and working data on your infrastructure. Cloud models receive context included in their requests, and connected services receive data sent by tools. Use local inference and deliberately controlled tools when data must remain on-premises.

Can several colleagues work with the same specialist?+

People with appropriate access can work with the same agent and its persistent context. Enable memory deliberately, configure crew knowledge, and review the role and permission model for your team.

Is Crewship ready for production?+

Crewship is open beta. Review release notes and feature documentation, test your workflows, and pin a release before using it for important work. Routines and integrations continue to evolve.

Can I install it on a phone?+

Installation requires a host with a supported container runtime. This mobile website lets you explore the platform and setup instructions; it does not install a native mobile application.

WELCOME ABOARD

One command
to get started.

Install Crewship on your computer or server. Then connect a supported container runtime and launch your first crew.

Full setup guide

Docker·Podman·Apple Containers

crewship — install

$brew install crewship-ai/tap/crewship

Then check your environment and start:crewship doctorcrewship start

Requires Homebrew and a supported container runtime.

View install script

BUILT IN THE OPEN

Follow what ships.

Release notes, improvements, and fixes. The product keeps moving; the changelog keeps the trail.

All hands onyour deck.

Welcome aboard.Let’s get to work.

Your people. Your agents.Your infrastructure.