[WIP]: This is still being worked on :D
Alright, today i feel a lot better, the fever is gone and the headache is now a lot smoother (even though its still present). I tried to cleanup some of my notes but it was hard to recall what happened 🤣 and so this was fruitless.
Hours on it: ~2am - ~8am
I started the day determined to make the tools that i needed the day before. I hacked together snippets from all over the internet and put together 3 very simple tools:
certnames: Parse a x509 PEM certificate and display Subject and/or DNS names with optional dns resolution of those alternative namescspparse: Connect to a given URL and parse the CSP server response header (if present) for URL schemes, hostnames and IP'ssniprobe: Connect to a given TLS port and parse the remote end certificate for DNS names. For every name perform a get request pinned at the original hosts' IP. This makes SNI happy and is very different from simply changing theHOST:header.
At the start of the day the pipeline looked something like this
I also followed through and organized the automation pipelines into more clearcut stages, which so far are like this:
initialization: Setting up any needed tools and files before the actual process startsinfo_gathering:passive: Information gathering stage with tools that dont access any of the targets' systems servicesinfo_gathering:active: Information gathering stage with tools that (at most) use a target system service but with minor "disturbance" (ieGET /or dns lookups using the targets dns servers etc)infogather_aggregation: This is an intermediate stage, it aggregates the information from all previous steps into anorunmila.dband makes all the artifacts from the previous steps available for a single download
This is how the pipeline looked like at the end of the day
The general plan for the day was
Since there was a plan, and i decided to follow it, it meant no new programs, just work on any of the past programs and try to keep notes of interesting findings.
I came across a lot of HubSpot instances this time around so i had to mess with it like any other application. Find their manuals and API references, check for endpoints and parameters. I was mostly trying to find endpoints that the application supports in hopes that some of them will be forgotten and unprotected.
My guess was right as i was able to find some endpoints for the usual user operations (register, signup, reset password, etc), but i didnt spend much time on them. Part of my process is to first gather as much information as possible and then try to make a plan of attack based on the available info.
I was so excited about the prospect of integrating the new tools as well as orunmila in a more meaningful way, into the mix. I was even more surprised,
when the tools worked flawlessly and got integrated into my methodology without any pains :D
orunmilafor holding all my findings and interesting logs and then allowing me to pipe their output at different other toolscspparsefor parsing csp policies and finding some extra domains that i was not able to find in any of my other scanssniprobefor probing SSL websites and listing domains that had interesting results (i loved error 530 from cloudflare 😃 )bbottried for its ability to do all in one kind of things but i was not able to observe any results yet, will have to look again at a later time.
I merged these two together for this day since i was mostly focusing on evaluating my findings and doing some manual testing.
I was able to cause a lot of application crash cases, which for me is a "good" sign that there is more to be found there.
I have eliminated amass and ffuf as they were taking way to long to complete at this time and they need to be provided with a more carefully crafted dataset to produce meaningful results in a timely manner. I decided it would be best, if these were added on followup steps and on better curated datasets (that i would do my self).
The community has very few tools outside the obvious ones (sub domains and website searches). Tools for more specialized and better DNS enumerations seem to be rare and tools for API's are mostly limited to zap and burp extensions, which make sense since these are far more complicated subjects in general.
Everything went much better than expected and i felt really pumped with my newly polished Gitlab pipeline. I had interesting results and was able to add a few more "knowledge" to my notes.
I kept on thinking that i am 6 days into my BugBounty journey and have started 5 tools already (steadily eyeballing on the 6th), the way i go i'll be lost in a sea of projects soon 🤣