Foundations: Law, Ethics and Scope · PTES: Pre-engagement
L3. Responsible Disclosure and Bug Bounties
Report a vulnerability through coordinated disclosure, read a bug bounty policy, and know what to do when you find a flaw outside any program.
Lesson 3 of 21 · About 20 minutes including the lab
By the end of this lesson you can
- Explain coordinated vulnerability disclosure and typical disclosure timelines.
- Find a vendor’s security contact through security.txt.
- Read a bug bounty policy for scope, safe harbor and forbidden actions.
- Handle an accidental discovery without making it worse.
Video transcript
You spot a flaw on a real site. What do you do next? Responsible disclosure means reporting a vulnerability privately to the owner and giving them time to fix it. Bug bounty programs add written rules, and sometimes rewards. Look for a security.txt file or a published disclosure policy to find the right contact. Send a clear private report with the steps to reproduce and the impact. Give the owner time to fix it, and coordinate before anything is published. A published policy tells you what testing is allowed and how to report. Without a policy, you have no permission to keep probing. Read the disclosure policy and its scope first. Share the details only with the security contact. Never demand payment in exchange for the details. Disclosure is part of the job. Practice the rules in the knowledge check.
Course outlineLesson 3 of 21
Lesson notes
Coordinated vulnerability disclosure (CVD) means you report a flaw privately to whoever can fix it, give them reasonable time to fix it, and only then discuss it publicly, ideally together. ISO/IEC 29147 describes how vendors should receive and publish vulnerability reports, and ISO/IEC 30111 describes how they handle them internally.
Timelines vary by organization. Google Project Zero, for example, publishes a 90-day disclosure policy, with extra time in some cases once a fix ships. The CERT Coordination Center uses a default of 45 days. Always follow the policy of the program you report to.
Many sites publish a security contact in a file at /.well-known/security.txt, defined by RFC 9116. It lists a Contact address and an Expires date, and can link to a policy, an encryption key and acknowledgments. Reading that file is an ordinary web request; it is not testing.
Bug bounty programs on platforms such as HackerOne, Bugcrowd and Intigriti publish exactly what is in scope, which vulnerability types they accept, and what is forbidden (commonly denial of service, social engineering, spam and accessing other users’ data beyond a minimal proof). Many programs include a safe harbor statement promising not to pursue researchers who follow the rules. Read the whole policy before you start, and re-read it when it changes.
If you stumble on a flaw outside any program, for example a page that shows other customers’ orders, stop. Do not keep exploring, download data, or test whether you can go further. Write down what you saw and when, report it through security.txt or the organization’s security contact, and if you cannot find one, ask your national CERT to help coordinate. Never demand payment in exchange for details; that can be treated as extortion.
A good report has a clear title, the affected asset, step-by-step reproduction, the impact you could demonstrate, and a suggested fix. Lesson 19 covers writing findings in detail.
Commands and expected output
Read a site’s security.txt (an ordinary page request)
curl -s https://securitytxt.org/.well-known/security.txtOutput
Contact: https://hackerone.com/ed
Expires: 2027-09-07T00:00:00.000Z
Acknowledgments: https://hackerone.com/ed/thanks
Preferred-Languages: en, fr, de
Canonical: https://securitytxt.org/.well-known/security.txt
Policy: https://hackerone.com/ed?type=team&view_policy=trueReal output fetched on 22 November 2026; the values will change over time. Contact is where reports go and Expires shows whether the file is maintained (both are required by RFC 9116). Policy links to the disclosure rules; Acknowledgments, Preferred-Languages and Canonical are optional.
Hands-on labs
Practice only on these legal targets: your own local lab or machines a platform assigns to you. Never use these techniques on systems you do not have written permission to test.
Hack The Box
Write up a retired Starting Point machineAfter you solve a Starting Point machine later in the course, write a short disclosure-style report: title, steps, impact and fix. Share write-ups only for retired content, as the platform’s rules require.
Practice task
Find the security.txt file for an organization you use every day, if it has one. Note the Contact and Policy fields. Then draft, but do not send, a three-paragraph disclosure email for a fictional IDOR issue in the practice booking app.
Key takeaways
- Report privately first, allow time to fix, then disclose in coordination.
- security.txt (RFC 9116) is the standard place to find a security contact.
- A bug bounty policy defines scope and forbidden actions; safe harbor only covers you if you follow it.
- On an accidental discovery: stop, document, report. Never ask for money for details.
Framework mapping
1. What is the core idea of coordinated vulnerability disclosure?
2. Where does RFC 9116 say to publish a security contact file?
3. A bug bounty policy forbids denial-of-service testing. You think a DoS finding would be valuable. What do you do?
4. While shopping you notice an order page shows another customer’s address. What is the right next step?