Getting access to 25000 employees details

Sahil Ahamad
Nov 8, 2017 · 3 min read

Hi guys,

I want to share one of my findings in a private program on HackerOne, which was — critical but straightforward one. During testing for that private program. I found an endpoint for Internal team management.

Internal Team management endpoint

After opening the endpoint (refer the Image above), the only thing running in my mind was “How about I check the directories.” Thus, I immediately utilized Dirsearch to brute force all the directories.

Here is the exciting output.

I renamed dirsearch to `dir` because I am lazy :v

Noticed? Anything?

It’s https://37.--.--.--/register :P

Upon opening the URL.

Yuss!!!! Registration page. 😮 anddd….

I tried to register with my details. And.. there was a configuration error. I was like…

I decided to register one more time with the same email and ended up with an error i.e.

“The email is already registered.”

okay, let’s go and log in.

So, I tried to log in with my registered credentials anddd…..

Successfully Logged in….

Admin management page.

All administrators name & email address was disclosed, I was even able to delete them.

Typical employee details pages

Disclosed details include Name, Email, Phone-No, Employee ID, Shifts, Reports, Salaries etc.

Typical Employee details (25k Records)

Sorry, but I needed to hide some details due to confidentiality issues. Some other critical data was disclosing too but don’t have permission to write further.

After verifying the issue, I quickly submitted the detailed report to the program via HackerOne. They validated and fixed the problem within a few hours.

They permanently fixed the issue by removing the public registration page from the endpoint.

After reporting the issue, I applied dirsearch on most of the critical endpoints belongs to them however no more endpoint was vulnerable to the same problem.


Report Submitted: 25–10–2017

Report Triaged: 25–10–2017

Initial 1300$ Awarded: 25–10–2017

Report closed as Resolved: 25–10–2017

Final 1200$ Awarded: 26–10–2017


I found this endpoint using Github issues conversations.

My recon process.

Sublister,knockpy,dnsresolver,dirsearch,bucket finder,massdns etc.

After reporting some low hanging issues, I go out and follow engineering/Security teams on Twitter and Github & look for anything interesting

I go through all the issues/Repositories companies engineering team created publicly on Github.

I read all blog posts by engineering and security team.

I check their DNS every month. Generally, companies stopped using a service and forgot to delete CNAMES pointing to service.

I use their services as the user and continue my recon processes,

I also use Burp Suite pro history tool to find exciting endpoints.

According to me, Recon is not a one time process it’s a continuous process.

If you like my blog posts and my work, Please consider checking out my “Buy me a coffee” page

Sahil Ahamad

Written by

Security Engineer @zomato | Information Security Researcher | Bug Bounty Hunter |