OOmnistu

Blog / Security

Security

security.txt explained: the small file that tells researchers how to reach you

One text file at /.well-known/security.txt gives security researchers a standard way to report vulnerabilities. The required fields, the optional ones, and where it fits for a small site.

OOmnistu
Reviewed 2026-11-20 · 3 min read

A security researcher just found a vulnerability in your website. Who do they email? If the honest answer is "no idea," the finding may never reach you - the securitytxt.org project puts it bluntly: when researchers lack a channel to disclose issues properly, security issues may be left unreported. security.txt is the standard answer: one plain-text file that says exactly how to report a vulnerability to you.

What goes in the file

Only two fields are required. Contact tells people how to reach you about security issues - each entry must start with mailto: for an email address, tel: for a phone number, or https:// for a web page such as a reporting form. Expires sets the date and time after which the file should be considered stale, in ISO 8601 format - so researchers know whether to trust it, and you have a reason to review it. The optional fields cover the rest of a disclosure setup: Policy links to the page describing what researchers may and may not do when testing and reporting, Encryption links to a key for secure communication, Acknowledgments links to a page thanking researchers, and Preferred-Languages, Canonical, Hiring and CSAF round out the less common cases.

Where it lives

For websites, the file goes at /.well-known/security.txt - the well-known path defined in RFC 8615 - with /security.txt at the root as a fallback when the well-known directory is not an option. It must be served as text/plain over HTTPS. If you want researchers to be confident the file is authentic rather than planted by an attacker, you can sign it with an OpenPGP cleartext signature, with the Canonical field pointing at its location.

Who uses it

The file format is an RFC - 9116, "A File Format to Aid in Security Vulnerability Disclosure" - and securitytxt.org lists adopters including Google, Facebook, GitHub and the UK government, with endorsements from the US CISA and the French, Italian, Dutch and Australian cyber authorities. For a small site the appeal is simpler than the company it keeps: the cost is one static file, and the payoff is that the next person who finds a hole knows where to write before they consider posting it publicly.

What it is not

security.txt answers one question - how do I tell you? It is not a bug bounty, and it does not by itself set the rules for testing. That is what the Policy field is for: it points researchers at your disclosure policy, so the file stays small even as the policy grows. Start with the two required fields, and add the rest when you have them.

Getting one up

Create the file with a Contact and an Expires date, publish it at /.well-known/security.txt over HTTPS, and set a reminder before the expiry date so the file never goes stale. That is the whole setup - the standard's job is to make the easy part universal so the hard part, fixing the bug, can start sooner.

Sources: securitytxt.org (field definitions, placement, FAQ, adoption and endorsements) and RFC 9116, "A File Format to Aid in Security Vulnerability Disclosure" (the standard itself).

Frequently asked questions

Is security.txt mandatory?

No - it is a voluntary convention standardized as RFC 9116. Large platforms use it and several national cyber authorities endorse it, but nobody requires it of a small site.

Which fields are required?

Two: Contact (how to reach you, with mailto:, tel: or https:// prefixes) and Expires (an ISO 8601 date after which the file should no longer be trusted). Everything else - Policy, Encryption, Acknowledgments and the rest - is optional.

Where exactly does the file go?

At /.well-known/security.txt, served as text/plain over HTTPS. If the .well-known directory is not possible, /security.txt at the site root is an accepted fallback - and you can publish both.

Will publishing an email address invite spam?

It does not have to be an email address. The Contact field accepts an https:// URI, so you can point researchers at a web form instead.