Identity infrastructure for agents
Your agent.
In the world.
Email, phone, browsers, containers, proxies, and Connections to the apps your team already uses. One identity for everything your agent does.
01One identity
Most agents act on borrowed credentials: your inbox, your browser, your keys. Lupin gives Your Agent an identity of its own, with every permission in your hands.
Not just a key.
A whole identity.
- 01 · Communicate
An identity people can reach.
An inbox and a phone number of its own. People write to Your Agent, and it answers from the same address.
- 02 · Act
Somewhere to do the work.
A browser that remembers, a container to run code, and a proxy for its own route to the web.
- 03 · Connect
Connected to your team’s tools.
Connections attach the apps your team already uses to the same identity, with only the access you allow.
02Agent passport
Meet Your Agent.
Everything it holds, under one identity. Choose a capability to see what it does, then run a local example.
Your Agent
Research & operations
- Agent ID
- agt_your-agent_01
- Project
- research
- Capabilities
- 6 available · 1 soon
CommunicateAvailable now
An inbox of its own.
Give Your Agent an address people can write to. It sends, receives, and follows each thread with the context to take the next step.
- Inboxes on lupin.sh or your own domain
- Threads, attachments, and scheduled drafts
- Webhooks for every email event
Permissionsread, send, reply
Waiting for an incoming message
03In action
One request, start to finish.
Follow Your Agent from an email in its inbox to a reply and a team update. Each step uses a different part of the same identity.
A request lands in its inbox.
Your Agent has its own address, so the request arrives with the full conversation attached.
your-agent@lupin.sh
Find our next offsite venue
Shortlist three venues for 12 people in Lisbon. Compare day rates and Wi-Fi, then send me your recommendation.
04Use cases
Different jobs. One foundation.
Start with the work, then give Your Agent only the capabilities that job needs.
From a watch list to a weekly brief.
Give a research agent what it needs to check sources on a schedule, work out what changed, and bring the findings back.
- Receive the watch list in its own inbox
- Visit each source in a persistent browser, through its proxy
- Compare the results in a container, email the brief, and save it to Notion
- Browsers
- Containers
- Proxies
- Connections
Illustrative setup. You choose the resources and permissions.
06Developers
A few lines.
A whole identity.
Start with the typed SDK, plain HTTPS, or the CLI. Or hand Lupin’s tools to the framework your agent already runs on.
Your agent’s inboxes, in one call.
Read your agent’s inboxes with a standard HTTPS request. This works with the limited key returned by autonomous registration.
Set LUPIN_API_KEY in your server environment. This example uses the built-in fetch API; no Lupin package installation is needed.
const response = await fetch( "https://api.lupin.sh/v0/emails/inboxes?limit=10", { headers: { Authorization: `Bearer ${process.env.LUPIN_API_KEY}`, }, },);if (!response.ok) throw new Error(`Lupin API: ${response.status}`);const { inboxes } = await response.json();for (const inbox of inboxes) console.log(inbox.email);Framework examples use Lupin’s SDK. Logos identify the tools shown; no partnership is implied.
07Pricing
Start free. Grow into it.
One plan for your agents’ inboxes, app actions, browsers, and sandboxes.
Paid overages require your consent and a monthly spending cap. Free has hard limits. Phone and proxy traffic are priced separately; product availability depends on provider configuration. View the console for active limits and usage.
08FAQ
Questions, answered.
The short version of how Lupin fits into the agents you build.
Explore the APIWhat is an agent identity?
It’s the resources and access an agent uses to act in the world: email, phone, a browser, a place to run code, a proxy, and connections to your team’s apps. Lupin brings them together as one identity, with permissions you control.
What can I use today?
Email, phone, proxies, browsers, containers, and Connections are available when their providers are configured. The console shows availability for your workspace. Virtual cards are coming soon.
What are Connections?
Connections are your agent’s authorized access to external apps. You choose the accounts it can use and the actions it can take. Connected apps belong to the agent’s identity; framework integrations are how developers build with Lupin.
Does Lupin work with my agent framework?
Lupin exposes a REST API, so you can connect from any framework or language that can make HTTPS requests. The OpenAPI specification and MCP endpoint describe the available operations. TypeScript SDK and CLI implementations are available to developers with source access.
Can I control what my agents can access?
Yes. Scope API keys to specific projects or inboxes, set permissions, and manage email allow and block lists. Each agent gets its own resources instead of sharing one set of credentials.
Can I use my own email domain?
Yes. Connect a custom domain, configure its DNS records, and create agent inboxes with addresses that belong to your brand.
How does pricing work?
Plans include email recipients, app actions, successful connections, browser runtime, and sandbox runtime. Start free, then choose Developer, Growth, Startup, or a contracted Enterprise plan. Paid overages require explicit consent and a monthly USD cap. Phone and proxy traffic are priced separately.
Intelligence meets
independence.
You build the agent. Lupin gives it a place in the world.