Research the web with Relay
Help Claude reach permitted websites that reject cloud datacenters, while keeping visibility and control on your computer.
Claude can research most public websites directly from a pod. Some sites reject all cloud datacenter traffic, require a real browser, or show useful content only after you sign in. Podbay Relay gives you a controlled way to help with those cases.
Start with the outcome
Ask Claude for the research you need, not a particular access method:
Compare these sources and prepare a brief with links. Use official APIs or public pages where possible. Tell me if a source requires my relay or signed-in access.
Claude first uses supported public routes and reports honestly when a site refuses access. It does not disguise bulk scraping or keep retrying a site that has said no.
When Relay helps
Relay is an optional program that runs on a computer you control. It lets selected pod traffic leave through your normal internet connection instead of the pod's datacenter connection.
This can help Claude:
- read a permitted public page that blocks datacenter addresses;
- use a real browser for a page that depends on JavaScript;
- read a site through your signed-in session after you explicitly enable that site;
- run a permitted project tool whose requests need your network identity.
Open Settings → Relay in Podbay to generate the setup command. Run that command on your own computer and leave Relay running while Claude needs it. One relay can serve all of your pods.
What you can see and control
The local Relay dashboard lets you:
- see which of your pods used it and which sites they reached;
- pause one pod without stopping the others;
- block a site;
- allow or revoke signed-in use for a specific site;
- stop Relay immediately.
Detailed activity stays on your computer. Podbay stores only coarse domain-level success or failure information needed to improve future access decisions.
What Relay does not expose
Your home files and private network are not available to the pod. Requests to local devices, private IP addresses, or LAN names are refused. Sites are opened signed out by default; your browser session is used only for domains you explicitly sign in to through Relay.
Traffic that is configured to require Relay fails when your computer is offline instead of silently falling back to the datacenter. Your computer must therefore be awake and Relay connected while that work runs.
Your network, your responsibility
Relay is not an anonymous proxy or a scraping bypass. Traffic looks like it came from your connection. Use it only for access you are allowed to make, and respect site terms, rate limits, and access controls.