The DetectDynamicJS Burp Extension provides an additional passive scanner that tries to find differing content in JavaScript files and aid in finding user/session data.
"A picture is worth a thousand words", please take a look.
Issues address in screenshots;
Duplicate reports
Second request also being sent as POST
Content-Length ignored
Status Code ignored
Content-Type ignored
I'm also working on it. If you had used 'inferredMimeType', I don't think it would have processed xml or html contents. Also, ignoring the '<' char if appeared first (as I had done previously), would reduce these false positives to some extent. The second request needs to be GET, ignoring the base request method. Content-length being ignored might be due to improper checking, and duplicate reports, ma be due to difference in parameters.
I'm also working on it, let me know if you have anything.
Well aware of all of them. Just didn't bother me too much yet. About
1: That one was on my list as next thing to implement. Patch coming soon (meaning about 2-3 weeks).
2: Is second on my list as things to implement. I'm not just going to drop POST, but see the other thread about this. If this can be done efficiently with Burp's API, I'm happy :) If the change from POST to GET (shifting the parameters to URL) returns a different answer, this should be a a finding. But not a Detected Dynamic JS, but exactly that behavior and criticality low.
4: There is a Content-Length check somewhere, will have to see why that one it.
5: Status code is addressed in the other comment on the first pull request.
6: Content-Type is totally unreliable. It almost has to be ignored. As example: Yahoo sometimes sends script files as application/json .
"A picture is worth a thousand words", please take a look.
Issues address in screenshots;
I'm also working on it. If you had used 'inferredMimeType', I don't think it would have processed xml or html contents. Also, ignoring the '<' char if appeared first (as I had done previously), would reduce these false positives to some extent. The second request needs to be GET, ignoring the base request method. Content-length being ignored might be due to improper checking, and duplicate reports, ma be due to difference in parameters.
I'm also working on it, let me know if you have anything.