Security.
How to report a security problem in anything we build, and how to test without putting anyone at risk.
Hub carries public health and safety information from agencies to the residents they serve. If you find a security problem in what we build, we want to hear about it. We don't pay rewards.
How to report
Email security@hub.inc. Please don't put details of an unfixed problem in a public GitHub issue, a support message or social media.
A good report has:
- the page, endpoint or app version affected
- steps we can follow to see the problem ourselves
- what an attacker could do with it, in your own words
- the test accounts you used
You can use AI tools. Check every step yourself before you send it, because we can only act on what we can reproduce.
Send us only your own data. If you saw someone else's, describe what you saw instead.
What we commit to
- A person will reply within 3 business days.
- We'll tell you within 5 business days whether the issue is valid and in scope, and within 2 business days for anything critical.
- We'll keep you updated until it's fixed, and you can check the fix for us if you'd like.
- We'll credit you publicly if you want us to.
What you can test
api.hub.inc
Our API. Every app talks to it, so most serious bugs end up here.ph.appandbeta.ph.app
The resident web app. Beta gets new releases first.- PHapp® for iOS and Android, version 4.0 and later
The resident app. That includes the Arizona and Bermuda editions at version 4.0 and later. portal.hub.incandportal.beta.hub.inc
Portal by Hub, the app agencies use to publish. Read the agency rules before you test it.account.hub.incandaccount.beta.hub.inc
Account settings.embed.ph.app
The widget agencies put on their own websites. Test it here or on a page of your own, never on an agency's site.www.ph.app,hub.inc,support.hub.inc,wehealth.organdwww.wehealth.org
Our websites. They're mostly static.
Anything not listed here is out of scope. If you're not sure, ask us first.
What you can't test
Please don't test these at all. If you come across a problem in one of them by accident, tell us. You're covered when you do.
- Our development and demo systems: anything under
dev.hub.inc,demo.hub.inc,dev.ph.appordemo.ph.app. - Our older system, which we're retiring. That means every host under
wehealth.orgexceptwehealth.organdwww.wehealth.org, pluslegacy.ph.app,portal-demo.hub.incand PHapp versions earlier than 4.0. - Our internal admin tool at
superadmin.hub.inc. Don't try to sign in, and don't guess passwords or codes for our staff. - Hosts on our domain that other companies run for us:
auth.hub.inc(Auth0),measure.hub.inc,links.hub.incandstream.hub.inc. The same goes for other companies' services we use, like Auth0, Twilio, SendGrid, Google Cloud, Firebase, HubSpot and Bunny. Report problems in their services to them. Our own settings on those services are ours, so a storage bucket or sign-in rule we set up is in scope through the targets above. - Our staff, the agencies we work with, and residents. That means no phishing, no social engineering, no calls and no physical access.
- Anything that would slow down or take down a service. Keep scripted or automated traffic under 5 requests a second.
What we don't treat as a vulnerability
Unless you show a real way to hurt someone, we don't treat these as vulnerabilities:
- missing security headers or cookie flags, or a content security policy you'd write differently
- email settings like SPF, DKIM and DMARC
- TLS versions or ciphers
- clickjacking on a page with nothing sensitive to click
- self-XSS, or anything that needs the victim to paste code into their own browser
- open redirects on their own
- software version numbers in a response
- rate limits, unless the missing limit lets someone send messages or run up our costs
- sending codes or invitations to an address you choose, where we already limit how many go out
- issues that need a rooted or jailbroken phone, or a browser that no longer gets updates
- reports from automated scanners that you haven't confirmed
- a way around our web application firewall on its own (it may block your tests, and a vulnerability you can exploit through it still counts)
- answers from our AI assistant that are wrong or off topic, unless you got it to reveal data or take an action it shouldn't
Rules for testing
People rely on Hub for emergency information. These rules keep them safe while you test.
Use your own accounts
Make up to 10 resident accounts with email addresses you own. If you need more, ask us. Don't use passwords, keys or data from a breach, a leak or another person, and don't try passwords on accounts that aren't yours.
Ask before you test the agency side
Agency access in Portal is given by hand. Never request access to a real agency, because those requests go to real staff at that agency. If you want to test what an agency can do, email security@hub.inc first and we'll talk about it.
Never try to read or change a real agency's posts, drafts, members or settings. If you think a real agency is exposed, stop and tell us.
Never send anything to a real person
That includes alerts, posts, push notifications, text messages, emails, invitations, messages to an agency or to our support team, and requests for agency access. The only people you may reach are you and the test accounts you made.
Text messages are switched off on our live system, so please don't try to turn them on.
Go easy on features that cost us money
Translation and our AI assistant cost us money every time they run. Keep your tests of them to a few dozen a day, and don't script them.
Stop at proof
If you can run code or get into a system, run something harmless, leave nothing behind, and don't use that access to reach anything else. If you find a password, key or token, don't use it. Tell us where you found it.
Stop when you reach real data
If you see anyone's personal information other than your own, stop, delete anything you saved, and tell us. You don't need to show many records. Show the problem with two accounts of your own and explain how it reaches others.
Mark your traffic
Send the header X-Hub-Research with your requests, set to the email address you report from.
Keep it private until it's fixed
Keep the details private until we've fixed the problem or 90 days after you reported it, whichever comes first. If we need longer, we'll tell you why before then and ask you for a new date. When you publish, leave out anyone's personal information and the names of the agencies involved.
Safe harbor
If you make a good faith effort to follow this policy, we'll treat your research as authorized. If you break one of these rules by accident and tell us, we'll still treat it that way.
We won't sue you or report you to the police, and we won't ask anyone else to, over good faith research under this policy. That includes claims under anti-hacking laws like the Computer Fraud and Abuse Act and anti-circumvention laws like the DMCA.
We waive the parts of our terms of service that would otherwise stop research done under this policy, for that research only.
If someone else takes legal action against you over research that followed this policy, we'll make it known that you acted with our permission.
We can only speak for Hub. We can't speak for the agencies we work with or for other companies, even ones whose services we use. If your finding exposed someone's personal information, the law may require us to tell them or the agency that serves them. We won't name you without asking.
The version of this page in force when you did the research is the one that applies.
Safe harbor doesn't cover extortion. That means asking for money or anything else in return for keeping a problem private or for telling us the details.
If you're not sure whether something is allowed, ask us at security@hub.inc before you do it.
Last updated November 21, 2026.
Found something?
Tell us what you found and how to see it. A person will reply within 3 business days.