Vulnerability disclosure policy template and security.txt builder
Updated
A vulnerability disclosure policy is a public page that says how to report a security problem and what the reporter can expect; a security.txt file is the small text file that points to it. No source fetched for this page says SOC 2 names either one, but the generated Vulnerability and Patch Management Policy promises both in item 4.7. This page builds the file to RFC 9116 and a one-page policy in Word or Markdown that quotes your own commitments. Built in your browser; no signup.
What the sources say
Three pages were fetched for this template on 17 November 2026: the specification itself and two vendors. None says the SOC 2 criteria name a disclosure policy, a security.txt file or a bug bounty. No audit-firm page on disclosure policies was found, so nothing below is a CPA firm’s view.
| Source | Who they are | What it says |
|---|---|---|
| RFC 9116 (E. Foudil and Y. Shafranovich, April 2022) (source) | An IETF document in the Informational category: "not an Internet Standards Track specification; it is published for informational purposes" | Defines the file format. The Contact field "MUST always be present"; the Expires field "MUST always be present and MUST NOT appear more than once". |
| HackerOne (HackerOne Team, 27 May 2022) (source) | Sells vulnerability disclosure and bug bounty programs | "Any website can set up a security.txt file in a few minutes, and it doesn’t require a full-fledged Vulnerability Disclosure Program (VDP) or bug bounty." |
| Drata help center (no author, 5 August 2026) (source) | Sells compliance automation software | "This standalone Responsible Disclosure Policy has been deprecated." For new customers: "we recommend incorporating responsible disclosure language into your Incident Response Plan or Vulnerability Management Policy". |
The HackerOne article describes security.txt as still under consideration by the IETF, although it is dated a month after RFC 9116; read it for the practical advice, and the RFC for the rules.
Why your own policy is the reason to do this
An examination tests a company against the controls it wrote down. These are the two statements about outside reports in the Vulnerability and Patch Management Policy, shown as they render for the sample company, Northwind Cloud Inc; with your answers saved in this browser the download quotes your own.
The commitment (Vulnerability and Patch Management Policy, 4.7)
Northwind Cloud Inc publishes a security contact at security@northwindcloud.example and a security.txt file, acknowledges good-faith reports within three business days, works with reporters to validate and fix issues, and does not pursue legal action against researchers who act in good faith, respect user privacy and avoid service disruption.
The routine (Vulnerability and Patch Management Policy, 5.7)
Disclosure handling. Reports to security@northwindcloud.example are acknowledged within three business days, reproduced and rated, tracked in the register and remediated within the standard timelines; the reporter is updated at triage and at resolution and credited with consent.
Every company that generated the policies has made these promises, whatever it answered in the generator. If the contact is published nowhere and the file does not exist, the policy and the website disagree, and that is the kind of gap the mock audit and a real one both look for. The dependency scanning guide applies the same reasoning to the fix deadlines in the same policy.
What RFC 9116 asks of the file
| Field | Required | What it is for |
|---|---|---|
Contact | Yes | How to report: an email address, a phone number or a web page. May appear more than once, in order of preference. |
Expires | Yes | The date and time after which the file is stale. Appears once. |
Acknowledgments | No | A page that recognizes people who reported issues. |
Canonical | No | The address or addresses where this file lives. |
Encryption | No | A link to a key for encrypted reports. The key itself does not go in the file. |
Hiring | No | A link to security job openings. |
Policy | No | A link to the vulnerability disclosure policy. |
Preferred-Languages | No | Languages reports may be written in. Appears once; with none listed, a reporter may assume English. |
The purposes in the table are summaries of the RFC’s field definitions, not quotations. The rules the builder enforces are the RFC’s own (RFC 9116, Apr 2022):
- Location. "organizations MUST place the "security.txt" file under the "/.well-known/" path".
- Transport and type. The file is reached over https, and "It MUST have a Content-Type of "text/plain"". Web addresses inside the file begin with
https://; the builder rejectshttp://. - Contacts are URIs. An email address is written with
mailto:and a phone number withtel:. The builder adds the prefix for you. - Expiry. "It is RECOMMENDED that the value of this field be less than a year into the future to avoid staleness." The builder warns past a year and refuses a date that has passed.
- Stale is worse than absent. "Not having a "security.txt" file may be preferable to having stale information in this file."
A finished file looks like this:
Contact: mailto:security@example.com
Expires: 2027-10-31T00:00:00Z
Preferred-Languages: en
Canonical: https://example.com/.well-known/security.txt
Policy: https://example.com/security
What is in the policy template
- Purpose
- Scope
- How to report
- What you can expect from us
- What we ask of you
- Rewards and recognition
- Changes to this page
- Filled from your generator answers: company, product, the Security Owner, the approver, the review cadence and the next review date.
- The reporting contact: the one you enter in the builder above, so the page and the file name the same address. Without it the template uses the security contact from your generator answers only when it is an email address, an https page or a phone number; a chat channel is never printed on a public page, and a placeholder is left instead.
- Quoted, not retyped: the acknowledgment time, the no-legal-action line and the handling routine come from items 4.7 and 5.7 of your policy. The template adds no time limit and no payment of its own, and a test fails if either item stops saying what the page relies on.
- Left as
[Fill in: ...]: the domains and applications in scope, anything else out of scope, how long you ask a reporter to wait before going public, whether you pay for reports, where credits are published, and the address of the page. - Not in it: legal review. The good-faith conditions are a starting text; have counsel read them if the wording matters to you.
Is a bug bounty part of this?
No fetched source ties one to SOC 2. HackerOne, which sells both, separates the two: "VDPs do not incentivize finders", while in a bounty program people are "paid based on bug severity or confirmed vulnerabilities". Its advice for an organization that "isn’t ready for an official program yet" is that "a great first step is to set up a dedicated security email address and a security.txt file" (HackerOne, 27 May 2022). That is what item 4.7 already commits you to, and nothing more.
One policy or two?
Drata’s help article says its standalone responsible disclosure policy "has been deprecated" and that the integrated approach "still addresses coordinated vulnerability disclosure expectations without requiring a separate policy document", then adds: "As audit expectations may vary, please consult with your auditor" (Drata, 5 Aug 2026). The generated set is already arranged that way: the rule sits inside the Vulnerability and Patch Management Policy. The download on this page is therefore not a 23rd policy. It is the public page that tells an outsider what the internal rule means for them. Drata’s sample text suggests its own response time; the template prints only the one in your policy.
Publishing both
- Create the address in the Contact field and check that mail to it reaches a person. Item 4.7 names the contact from your generator answers; if that is a chat channel, give outsiders an email address or a web form instead, because the file needs an email address, a phone number or an https page.
- Download the policy, replace every
[Fill in: ...], have the approver sign off and publish it on your site. A trust page is a natural place to link it from. - Put the policy’s address in the builder, download
security.txtand publish it under/.well-known/. Open the address in a browser and check that it loads over https as plain text. - Add the renewal date to the calendar, with the policy review.
- Tell your penetration tester about the page: the penetration test scope and the disclosure policy should name the same systems.
What to keep
- The published file and the policy page, with the date each was last changed.
- For each report: the message, the date it was acknowledged, the rating, the register entry and the date it was resolved. Item 5.7 says reports are "tracked in the register", the same register as every other finding, on the fix timelines in item 4.3.
- The updates sent to the reporter at triage and at resolution, which 5.7 also promises.
- A period with no reports: the file, the page and a note that the inbox was checked. See the evidence checklist for where this sits among the other records.
How it relates to the policies and the Audit Kit
The Vulnerability and Patch Management Policy holds the rule; this template is the public page and the file that the rule refers to. It is a free companion and is not one of the 22 policies or one of the files in the Audit Kit, so the downloads on this page are the whole thing. How often to scan is covered in the vulnerability scanning guide, and more free documents are on the templates page.
Start in the generator, then come back here to build the file and the policy.
Frequently asked questions
- Does SOC 2 require a vulnerability disclosure policy or a security.txt file?
- None of the three sources fetched for this page says the criteria name either one, and no audit-firm page on the question was found. What is certain is narrower: if you adopted the Policyseed policies, item 4.7 of your Vulnerability and Patch Management Policy says the company publishes a security contact and a security.txt file, so both are now your own written rule. Ask your CPA firm how it tests that item.
- Is a bug bounty required for SOC 2?
- No fetched source says so. HackerOne, which sells bounty programs, writes that a security.txt file does not require a full vulnerability disclosure program or a bug bounty, and describes a disclosure program as one that does not pay finders. The template's rewards section is a placeholder: state plainly whether you pay, and do not promise a payment you have no budget for.
- Which security.txt fields are required?
- Two. RFC 9116 says Contact must always be present and Expires must always be present and appear only once. Acknowledgments, Canonical, Encryption, Hiring, Policy and Preferred-Languages are optional.
- Where does the security.txt file go?
- Under /.well-known/ on the site it describes, for example https://example.com/.well-known/security.txt. RFC 9116 says the file must be reached over https and served as plain text. A copy at the top level of the site is allowed for older tools, but when both exist the one under /.well-known/ is the one that counts.
- How long should the expiry date be?
- RFC 9116 recommends a date less than a year ahead, so stale contact details do not stay published. The builder suggests 350 days and names a day 30 days earlier for a calendar reminder; both numbers are this page's choice, not the RFC's. The RFC also says that having no file may be preferable to having a stale one.
- Do I need to sign the file?
- No. RFC 9116 recommends an OpenPGP cleartext signature and, when the file is signed, a Canonical field so the signature also covers where the file lives. Neither is required. The builder does not sign anything; if you sign, do it after the last edit, because changing the expiry date invalidates the signature.
- What if nobody ever reports anything?
- No fetched source covers how an empty period is examined. This page's reading of your own policy: item 4.7 is about publishing the contact and the file, which you can show at any time, and item 5.7 describes what happens when a report arrives. Keep the published file, the page and a note that the inbox was checked; a period with no reports is a fact to state, not a gap to fill.
- Is anything uploaded?
- No. The security.txt file and the Word and Markdown policy are built on your device. The generator keeps your answers in this browser's local storage; with no saved answers the policy uses the sample company, Northwind Cloud Inc.
Policyseed provides governance policy templates and AI tailoring. It is not legal advice and not a compliance guarantee. Management adopts the policies; the CPA firm performs the SOC 2 examination.