
Introduction
In Bugbounties “If you are not first , then you are last” there is no such thing as silver or a bronze medal , Recon plays a very crucial part and if you can detect/Identify a newly added asset earlier than others then the chances of you Finding/Reporting a security flaw on that asset and getting rewarded for the same are higher than others.
Personally I am monitoring CT logs for domains/subdomains for quite a long time now and it gave me a lot of successful results , The inspiration behind this was “Sublert : By yassineaboukir” which checks crt.sh for subdomains and can be executed periodically , However I am using somewhat different approach and instead of looking into crt.sh periodically, I am extracting domains from Live CT log feeds , So chances of me finding a new asset earlier is higher as compared to others.
Workflow
- Monitoring Real Time CT log feed and extracting the domain names from that feed
- Matching the extracted subdomains/domains against the domains/Keywords to be matched
- Sending a Slack notification if a domain name matches
Requirements :
- A VPS (UNIX up and running)
- Python 3x (Tested with Python 3.6.9)
- Slack Workspace (optional)
Setup
I am assuming that you have already done with your setup of slack workspace .
Now Create a channel named “subdomain-monitor” and set up a incoming webhook
Enabling Slack Notifications :
Edit config.yaml
file and paste your slack webhook URL there , It should look something like this

Keywords and domains to match :
You can specify keywords and domains to match in domains.yaml
file , You can specify names
For Matching subdomains :

Note : Notice that preceding dot [ . ]
Lets take “.facebook.com” as example , domains extracted from Real time CT logs will be matched against the word “.facebook.com” , if matched they will be logged in our output file (found-domains.log) . The thing to note here is , It will give some false positives like “test.facebook.com.test.com” , “example.facebook.company” but we can filter out them later on by using use regex magic
For Matching domains/subdomains with specific keywords :
Lets assume that you want to monitor and log domains/subdomains that are having word “hackerone” in them , then our domains.yaml file will look something like this

Now all the extracted domains/subdomains that are having word “hackerone” in them will be matched and logged (and a slack notification will be sent to you for the same)
Okay we are done with our initial setup , Lets install the required dependencies and run our tool
$ pip3 install -r requirements.txt
$ python3 certeagle.py

Matched domains will look like this :

Slack Notifications will look like this :

Output files :
The program will keep on running all the matched domains will be saved under output directory in found-domains.log file

Strict Warning : Do not monitor assets of any organisation without prior consent
Inspiration
Contact
Shoot my DM : @0xAsm0d3us
Leave a Reply