Joining a team on Hamster

Welcome — someone on your team invited you to Hamster. This guide gets you oriented in about 10 minutes.

This guide:

• Is tailored for: individual contributors joining an existing Hamster workspace. Product managers, engineers, engineering managers, designers, and business stakeholders.
• And covers: what Hamster is, how to set up your account, what to do first based on your role, and where to go next.

Setting up Hamster for your team from scratch? See the admin guides instead: Small teams · Startups & mid-size · Enterprise & scaling.

What Hamster is

Hamster is the place your team's organisational understanding lives — what your products are, what's being built next, how your team works — in a structured English layer that AI can read. It sits on top of GitHub and Linear or Jira and the rest of your stack; it doesn't replace them.

If you want the full mental model, see the concepts overview. The short version:

• Blueprints capture state — what your product is today.
• Briefs capture change — what's being built next, refined against the Blueprint.
• Methods capture ways of working — how AI helps your team consistently.
• Your Context Graph connects all three to your existing tools.

For the conceptual model behind these artefacts, see Concepts — Delivery, Discovery, Direction, and Knowledge.

You'll mostly interact with briefs day-to-day. Reading the four concept pages takes about 15 minutes and saves a lot of "what does this thing actually do?" later.

Set up your account

1. Accept the invite. Click the link in your email and sign up using the same email address.
2. Step through onboarding. It's a couple of minutes. Hamster pulls your team's identity from your company URL, so most fields auto-populate.
3. Install the Slack bot. Most teams use Slack as the primary notification surface. Notifications, alignment requests, and brief side panels all land where you already work.
4. Optional: set up Voice. If you want hands-free brief writing while you walk or commute.

↳ Doc: Onboarding

What to do first, by role

Hamster's surface area is wide; what's relevant on day one depends on what you do on the team. Pick the section below that fits.

If you're a Product Manager

Open a chat and describe a small piece of work you'd normally write a brief for. Let the AI help you draft it — answer its clarifying questions, attach any relevant context (Figma frames, customer call clips, screenshots, URLs), and let it produce a brief.

Once the brief looks right, generate a plan. Vote alignment with a teammate (the brief moves to "Aligned" status). Click Deliver. A PR appears in GitHub.

That's the loop end-to-end. Read Briefs and Plans when you have 20 minutes for depth.

If you're an Engineer or Engineering Manager

You stay in your IDE. Hamster's job is to give your existing AI coding tool — Claude Code, Cursor, Codex, whichever you already use — full access to your team's Context Graph, the brief you're shipping, the relevant blueprints, and the methods. You ship the way you already ship. The brief just shows up with everything the agent needs to do the work.

1. Install the Hamster CLI. The CLI gives you the brief surface in your terminal — sync skills, run slash commands, trigger deliveries, scope contexts.
2. Wire up the MCP server. This is the part that matters. The MCP server exposes Hamster's Context Graph, briefs, blueprints, and methods to your AI assistant. From Claude Code or Cursor, you can ask it to ship a specific brief by name and the agent already has every piece of context it needs.
3. Skim the Methods Library. The methods are the playbook the AI reads from. If your team has forked it, knowing what's customized is useful — but you don't have to memorise it.
4. Cloud Agents are optional. Cloud Agents are how your admin gets briefs delivered without an engineer's laptop in the loop — useful for deliveries kicked off by a PM or for background work. As an engineer, you don't have to use them. Most engineers ship from their IDE and that's intentional.

If you're a Designer

You don't need to write briefs. You attach to them.

1. Connect Figma to your account. Workspace-level connection but each individual user links their own Figma account.
2. Find a brief your PM is drafting. Attach the relevant Figma frame or .fig file. The AI agent reads it as design context for the plan and delivery.
3. Watch the activity timeline on briefs you care about. It shows scope changes — including AI-driven ones — as they happen, so you don't get blindsided when a PR shows up that misses the design.

If you're a Business Stakeholder (Exec, Sales, CS, Ops)

You probably won't open Hamster directly day-to-day. Live in Slack.

1. Make sure the Slack bot is in your DMs. Notifications and brief side panels land here.
2. Drop customer signal directly into Slack. Use @Hamster to mention the bot in any channel; paste a customer quote, a Gong clip URL, or a support ticket. The bot routes it to the right brief.
3. Use slash commands to ask "what's shipping this week?" or "status of [initiative]?" without leaving Slack.
4. Forward public share links to customers, advisors, or board members. They can read the brief without a Hamster login.

Tips for a fast first week

• Don't try to learn all of Hamster at once. The brief surface plus Slack covers about 80% of day-to-day value. Adopt initiatives, routines, and methods later.
• Open chat first, not Notion. Hamster's chat is the brief writer. Pasting a finished doc loses the back-and-forth that makes briefs better.
• The AI reads everything you attach. Figma frames, Gong clips, screenshots, support tickets, URLs. More grounding context = better briefs and better deliveries.
• The activity timeline is your friend. If you're not sure what changed on a brief, look there before pinging Slack.

Where to go next

• The Hamster Studio reference — every feature, end-to-end.
• The admin guide for your team's stage, if you're curious about the bigger picture: Small teams · Startups & mid-size · Enterprise & scaling.