Customer Support

How to Reply to a Security Vulnerability Report

A practical guide for indie developers and small teams on replying to security vulnerability reports. Covers acknowledging receipt, verifying the issue, setting timelines, spotting beg bounties, and crediting reporters, with reply templates.

SupportMe••10 min read

When someone emails to say they've found a security hole in your product, a good reply follows four steps. Confirm you got the report, give a timeframe for your next update, ask for whatever you need to verify it, and keep the conversation private until a fix ships. Don't threaten the reporter, don't argue about severity in your first message, and don't pay anyone before you've seen what they found.

The rest of this guide covers each stage in more detail. It includes reply templates you can adapt and advice on handling the awkward cases, such as researchers who ask for money up front.

Why the first reply matters so much

A reporter who hears nothing back can't tell whether you're working on the issue or ignoring it. Silence is a common reason reporters go public before a fix is ready.

The OWASP Vulnerability Disclosure Cheat Sheet calls the communication phase "often one of the hardest points" of the whole process. It recommends regular status updates even when you have no firm dates, because they show the issue hasn't been forgotten.

CERT/CC lists "Avoid Surprise" as one of its core principles in The CERT Guide to Coordinated Vulnerability Disclosure. In practice, that means you and the reporter should both know what's coming next.

Step 1: Acknowledge receipt quickly

Your first reply doesn't need to say whether the bug is real. It just needs to confirm the report arrived and tell the reporter when they'll hear from you again.

There's no single legal standard for response time, but published policies give useful benchmarks. CISA's vulnerability disclosure policy template for US federal agencies commits to acknowledging reports "within 3 business days." For a small team, a promise of a few business days is realistic and easy to keep.

Template: first acknowledgement

Hi [Name],

>

Thanks for reporting this to us. We've received your report about [short description, e.g. "the password reset flow"] and we're looking into it.

>

We'll send you an update by [date]. Until then, please keep the details private while we investigate.

>

If you have more information, such as reproduction steps, affected URLs or versions, or a proof of concept, you can reply to this email directly. [Optional: Our PGP key is at ...]

>

Thanks,
[Your name]

Keep it short. Don't confirm or deny the vulnerability yet, and don't promise a fix date you haven't checked.

Step 2: Verify the report before you judge it

Next, work out whether the issue is real, how severe it is, and what it affects. OWASP suggests this order: ask for clarification if needed, confirm the issue, ask the reporter to retest the fix if that's useful, then confirm the fix.

If the report is vague, ask specific questions. Generic requests for "more info" slow everyone down. Many of the techniques in how to ask for better bug reproduction steps also work for security reports. Ask for the exact request, the account role used, the version or environment, and what data was exposed.

Some practical checks while you verify:

  • Reproduce the issue in a test environment, not in production, where you can.
  • Check your logs to see whether anyone other than the reporter has exploited the issue. If they have, you may be dealing with an incident and not just a vulnerability.
  • Don't let severity labels decide for you. Judge the impact yourself, whether the report says "critical" or "minor."
  • Keep the discussion in a private channel. Don't post it in a public issue tracker or forum.

If you maintain an open source project on GitHub, repository security advisories let you discuss the issue privately and fix it in a temporary private fork. You can also credit the reporter there and request a CVE ID. GitHub is a CVE Numbering Authority and says it usually reviews CVE requests within 72 hours.

Step 3: Send a clear status update

Once you know what you're dealing with, tell the reporter. Your update will usually take one of three forms.

Template: confirmed

Hi [Name],

>

We've confirmed the issue you reported. [One sentence on impact, e.g. "It allowed authenticated users to read other accounts' invoices."] We're working on a fix and expect to ship it by [date or range].

>

We'll let you know when it's deployed. If you're willing to retest at that point, that would be very helpful.

>

Would you like to be credited when we publish the fix? If so, tell us the name or handle you'd like us to use.

Template: can't reproduce yet

Hi [Name],

>

We haven't been able to reproduce this yet. We tried [what you tried] on [version/environment]. Could you share [specific missing detail]?

>

We're keeping the report open and will follow up by [date].

Template: not a vulnerability, or accepted risk

Hi [Name],

>

Thanks again for the report. After reviewing it, we don't consider this a security vulnerability because [specific reason, e.g. "the endpoint only returns data the user already has access to"].

>

If you think we've missed something, please let us know. We're happy to take another look.

Even when you decline, explain your reasoning and leave the door open. The reporter may know something you don't.

If you're behind schedule, say so before the date passes. A late update that explains the delay is better than no update. How to use reply time estimates covers how to set expectations you can keep.

Step 4: Agree on disclosure timing

Many researchers expect to publish their findings at some point. Agree on a date early so neither side is surprised.

Public deadlines from well-known programs show what "normal" looks like:

  • Google Project Zero gives vendors 90 days from notification to fix a bug. If the fix ships in that window, details are published 30 days after the patch. For bugs already being exploited in the wild, the deadline drops to 7 days.
  • CERT/CC begins planning to publish a public Vulnerability Note 45 days after first attempted contact with the vendor, unless a different date is negotiated.

These are those organizations' own policies. Individual researchers may follow different timelines. If you need more time, ask early and explain why. Most reasonable requests go better when they come with a concrete plan.

Step 5: Fix, confirm, and give credit

When the fix is live:

  1. Tell the reporter and ask them to retest if they offered to.
  2. Credit them in your changelog, advisory, or a security acknowledgements page, if they want to be named.
  3. For software that customers install or run, think about publishing an advisory.

OWASP recommends that an advisory include a summary and impact, affected versions, fixed versions, any workarounds, and a CVE ID if one was assigned. It also notes that disclosure "does not make the vendor look bad."

About rewards: you don't need a paid bug bounty to treat reporters well. OWASP suggests options such as service credit, swag, or a "hall of fame" listing for reporters who act professionally. It also cautions that formal bounty programs suit organizations that already have "a mature vulnerability disclosure process."

How to handle "beg bounty" emails

Small teams often receive unsolicited emails that claim to have found a vulnerability and ask for payment, sometimes before sharing any details. Sophos researcher Chester Wisniewski named these "beg bounties". He found the reports often come from automated scans of basic misconfigurations, such as a missing DMARC record, with fees requested from $150 to $2,000 per issue.

Here's how to respond:

  • Don't pay for details up front. OWASP notes that some of these requests may be legitimate but many are scams. It suggests asking the sender to report through a mediated bug bounty platform, which scammers tend to avoid.
  • Check the claim yourself anyway. A pushy sender can still point to a real problem. Missing email authentication records, for example, are easy to confirm and fix.
  • Escalate threats. If the message threatens to publish or exploit the issue unless you pay, treat it as a security and legal matter, not a support ticket.

Template: payment requested before disclosure

Hi,

>

Thanks for getting in touch. We review all security reports, but we don't pay for vulnerability details before seeing them. If you share the details of the issue, we'll assess it and follow up.

>

[Optional: We may offer recognition for valid reports at our discretion, as described in our security policy at ...]

Make the next report easier to receive

Once you've handled one report, a little setup will make the next one go more smoothly.

  • Publish a security.txt file. RFC 9116 defines a plain-text file at /.well-known/security.txt that tells researchers how to contact you. It needs a Contact field and an Expires field, and the RFC recommends setting the expiry date less than a year ahead.
  • Write a short disclosure policy. CISA's template is a good starting point. It covers scope, safe harbor for good-faith research, and response commitments. CISA also says it "will not recommend or pursue legal action" against researchers acting in good faith under its policy. Wording like that helps researchers feel safe reporting to you.
  • Train whoever reads the inbox. OWASP recommends that frontline staff learn to recognize security reports and know who to escalate them to. On a two-person team, that might simply mean agreeing who owns them.
  • Add security reports to your reply checklist. A support reply checklist can include "acknowledge within X days," "move to a private channel," and "set the next update date."

A vulnerability report from a researcher isn't the same as a customer asking about your security practices. For that kind of question, see how to answer security questions in 10 minutes.

A note on AI drafts for security replies

AI drafting tools can help with the routine parts of these replies, such as the acknowledgement and status-update templates. Every security reply still needs careful human review. A wrong detail can confirm a vulnerability before you meant to, promise a date you can't keep, or reveal technical details that should stay private.

SupportMe is built around human review: nothing is sent without your explicit approval. Even with that safeguard, read security replies line by line. Decide whether exploit details should go into any third-party tool at all. For help deciding when to draft by hand, see AI draft or manual reply? A support decision guide.

Conclusion

A good reply to a security vulnerability report is quick, private, specific, and honest about timing. Acknowledge the report within a few days, verify it carefully, send updates before your stated dates, agree on disclosure timing, and credit the reporter once the fix is live. Be polite but firm with payment demands, and still check whether the issue they point to is real. A security.txt file and a short disclosure policy make each new report easier to handle.

References

Tags

security vulnerability reportvulnerability disclosurecoordinated vulnerability disclosuresecurity researcher replybeg bountysecurity.txtresponsible disclosure email

Related posts