There is no public inbox listed on this site right now, so GitHub is the best way to reach the project.
Use GitHub for now
Contact us
If you need to reach the Relay project, keep it direct.
If something is broken, report the broken thing. If you are trying to figure out whether Relay fits your workflow, send that context. The goal here is not a fluffy contact page. It is to make it clear where to go and what helps.
Best first path
Broken deploys belong in GitHub issues.
Setup questions usually go faster if you compare against the docs first.
No public email is listed on the site yet.
How to reach us
Relay is still a technical project more than a polished support organization, so the contact options are practical on purpose. If you need help, the best path is usually the one that keeps context attached to the problem.
That means GitHub for bugs and product discussion, the repo for code and pull requests, and email only if a public inbox is actually configured for this site.
If you are writing in about a deploy problem, a short, concrete report is much more useful than a long general description. Command, runtime, deploy id, and log output usually matter more than anything else.
What are you contacting us about?
Most conversations fall into one of these buckets.
Keeping the reason clear makes the response clearer too.
A deploy is broken and you want help
Send the deploy id, the relevant log section, the runtime engine, and what you expected to happen.
You are still setting it up
Start with the docs or quickstart first, then reach out if the setup still does not match what you are seeing.
You want to know if Relay fits
Share the shape of your setup: app type, hosting model, whether you need previews, and how much of the stack you want to keep self-hosted.
If you are reporting a bug
The helpful version is the one with the real details.
You do not need to write a perfect report. You do need to include the pieces that make the problem reproducible or at least understandable.
Which command you ran.
Whether the app is on Docker or Station.
Host OS and architecture.
The deploy id and preview URL, if one was created.
The log section around the failure.
Redacted .relay.json, relay.config.json, or relay.json if config seems related.
Before you send logs
Redact tokens, cookies, private hostnames, and secrets.
Keep the command, buildpack, runtime messages, and deploy id intact.
Those details are usually the shortest path to understanding what actually went wrong.
Before opening an issue
If it looks like a setup mismatch rather than a bug, compare against the docs and quickstart first. That will usually answer the question faster than a round-trip.