DanMcInerney/xsscrapy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

added http header xss testing

Browse Source
This commit is contained in:
DanMcInerney
2014-07-11 05:13:36 -04:00
parent 882bb5ffab
commit 530c175194
3 changed files with 150 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
xsscrapy/items.py
View File

@@ -6,5 +6,6 @@
from scrapy.item import Item, Field
class vuln(Item):
xss_type = Field()
vuln_url = Field()
error = Field()

3
xsscrapy/settings.py
View File

@@ -29,3 +29,6 @@ ITEM_PIPELINES = {'xsscrapy.pipelines.XSS_pipeline':100} # Look into what the 10
FEED_FORMAT = 'csv'
FEED_URI = 'vulnerable-urls.txt'
#COOKIES_DEBUG = True
# Test for injection via headers
#DEFAULT_REQUEST_HEADERS = {'Referer': '9zqjx', 'User-Agent':'9zqjx'}

216
xsscrapy/spiders/xss_spider.py
View File

@@ -27,7 +27,7 @@ class XSSspider(CrawlSpider):
#allowed_domains = ['']
#start_urls = ['']
#rules = (Rule(SgmlLinkExtractor(deny=('logout')), callback='parse_resp', follow=True), )
#rules = (Rule(SgmlLinkExtractor(deny=('logout')), callback='parse_resp', follow=True), ) # prevent spider from hitting logout links
rules = (Rule(SgmlLinkExtractor(), callback='parse_resp', follow=True), )
def __init__(self, *args, **kwargs):
@@ -46,6 +46,9 @@ class XSSspider(CrawlSpider):
self.login_user = kwargs.get('user')
self.login_pass = kwargs.get('pw')
def parse_start_url(self, response):
return list(self.parse_resp(response))
#### Handle logging in if username and password are given as arguments ####
def start_requests(self):
if self.login_user and self.login_pass:
@@ -56,7 +59,11 @@ class XSSspider(CrawlSpider):
def login(self, response):
args, url, method = fill_login_form(response.url, response.body, self.login_user, self.login_pass)
self.log('Logging in...')
return FormRequest(url, method=method, formdata=args, callback=self.confirm_login, dont_filter=True)
return FormRequest(url,
method=method,
formdata=args,
callback=self.confirm_login,
dont_filter=True)
def confirm_login(self, response):
if self.login_user.lower() in response.body.lower():
@@ -68,7 +75,6 @@ class XSSspider(CrawlSpider):
############################################################################
def parse_resp(self, response):
payloaded_urls = None
resp_url = response.url
body = response.body
doc = lxml.html.fromstring(body, base_url=resp_url)
@@ -76,6 +82,16 @@ class XSSspider(CrawlSpider):
reqs = []
payload = self.test_str
# Edit a few select headers with injection string and resend request
header_reqs = [Request(resp_url,
headers={'Referer':self.test_str,
'User-Agent':self.test_str},
callback=self.make_header_reqs,
meta={'type':'headers',
'payload':self.test_str},
dont_filter=True)]
reqs += header_reqs
# Fill out forms with xss strings
if forms:
form_reqs = self.make_form_reqs(forms, resp_url, payload, injections=None, quote_enclosure=None)
@@ -85,75 +101,122 @@ class XSSspider(CrawlSpider):
if '=' in resp_url:
payloaded_urls = self.makeURLs(resp_url, payload)
if payloaded_urls:
payloaded_reqs = [Request(url, callback=self.xss_reqs, meta={'payload':self.test_str, 'orig_url':url}) for url in payloaded_urls] # Meta is the payload
payloaded_reqs = [Request(url,
callback=self.url_xss_reqs,
meta={'payload':self.test_str,
'resp_url':url})
for url in payloaded_urls] # Meta is the payload
reqs += payloaded_reqs
# Each Request here will be given a specific callback relative to whether it was URL variables or form inputs that were XSS payloaded
return reqs
def make_form_reqs(self, forms, orig_url, payloads, injections, quote_enclosure):
def make_header_reqs(self, response):
''' Creates a duplicate request for each link found with injected referer and UA headers '''
resp_url = response.url
body = response.body
doc = lxml.html.fromstring(body)
quote_enclosure = self.single_or_double_quote(body)
payload = response.meta['payload']
reqs = []
if payload in body:
injections = self.inj_points(payload, doc)
payloads = self.xss_str_generator(injections, quote_enclosure)
reqs = [Request(resp_url,
headers={'Referer':pload, 'User-Agent':pload},
callback=self.xss_chars_finder,
meta={'payload':pload,
'resp_url':url,
'quote':quote_enclosure,
'type':'headers'},
dont_filter=True)
for pload in payloads]
for r in reqs:
print r
return reqs
def make_form_reqs(self, forms, resp_url, payloads, injections, quote_enclosure):
''' Logic: Get forms, find injectable input values, confirm at least one value has been injected,
confirm that value + url + POST/GET has not been made into a request before, finally send the request '''
reqs = []
vals_urls_meths = []
url = None
test_or_payload = 'payload'
if type(payloads) == str:
if type(payloads) == str: # type(payloads) is not a list
test_or_payload = 'test'
payloads = [payloads]
for payload in payloads:
for form in forms:
values, url, method = self.fill_form(orig_url, form, payload)
# If the form is a GET not a POST, then just send it to the normal queue
# if method == "GET":
# print 'SENDING THIS ONE BACK DOWN THE PIPELINE!!!!'
# print values
# req = [FormRequest(url, formdata=values, method=method, callback=self.parse_resp)]
# print req
# reqs += req
# continue
if url == None:
self.log('No form action URL found')
continue
if len(values) == 0:
self.log('No values changed, aborting this form test')
continue
# Sometimes lxml doesn't read the form.action right
if '://' not in url:
self.log('Form URL contains no scheme: %s' % url)
continue
# Make sure at least one value has been injected
if not self.injected_val_confirmed(values, payload):
self.log('Form contains no injected values: %s' % orig_url)
values, url, method = self.fill_form(resp_url, form, payload)
url = self.check_form_validity(values, url, payload, resp_url)
if not url:
continue
if not self.dupe_form(values, url, method, payload):
# Make the payloaded requests, dont_filter = True because scrapy treats url encoded data as == to nonurl encoded
if test_or_payload == 'test':
req = FormRequest(url, callback=self.form_cb, formdata=values, method=method, meta={'payload':payload, 'injections':injections, 'orig_url':orig_url}, dont_filter = True)
cb = self.form_cb
elif test_or_payload == 'payload':
req = FormRequest(url, callback=self.xss_chars_finder, formdata=values, method=method, meta={'payload':payload, 'injections':injections, 'quote':quote_enclosure, 'orig_url':orig_url}, dont_filter = True)
cb = self.xss_chars_finder
req = FormRequest(url,
callback=cb,
formdata=values,
method=method,
meta={'payload':payload,
'injections':injections,
'quote':quote_enclosure,
'resp_url':resp_url,
'type':'form'},
dont_filter = True)
self.log('Created request for possibly vulnerable form')
reqs.append(req)
return reqs
def check_form_validity(self, values, url, payload, resp_url):
''' Make sure the form action url and values are valid/exist '''
# Make sure there are values to even change
if len(values) == 0:
self.log('No values changed, aborting this form test')
return
# Make sure at least one value has been injected
if not self.injected_val_confirmed(values, payload):
self.log('Form contains no injected values: %s' % resp_url)
return
if url == None:
self.log('No form action URL found')
return
# Sometimes lxml doesn't read the form.action right
if '://' not in url:
self.log('Form URL contains no scheme, attempting to put together a working form submissions URL')
proc_url = self.url_processor(resp_url)
url = proc_url[1]+proc_url[0]+url
return url
def fill_form(self, url, form, payload):
if form.inputs:
for i in form.inputs:
if not self.input_filter(i):
continue
if 'InputElement' in str(type(i)): # gotta be a better way to do this
if i.name:
# Keep submit input value the same
if i.type == 'submit':
form.fields[i.name] = i.value
continue
if 'InputElement' in str(type(i)): # gotta be a better way to do this
if i.name:
form.fields[i.name] = payload
values = form.form_values()
return values, form.action or form.base_url, form.method
def injected_val_confirmed(self, values, payload):
@@ -190,7 +253,7 @@ class XSSspider(CrawlSpider):
return True
def form_cb(self, response):
orig_url = response.meta['orig_url']
orig_url = response.meta['resp_url']
payload = response.meta['payload']
body = response.body
doc = lxml.html.fromstring(body)
@@ -201,12 +264,6 @@ class XSSspider(CrawlSpider):
injections = self.inj_points(payload, doc)
payloads = self.xss_str_generator(injections, quote_enclosure)
#print ''
for i in injections:
print i
#for p in payloads:
# print p
form_reqs = self.make_form_reqs(forms, orig_url, payloads, injections, quote_enclosure)
for r in form_reqs:
print ' FORM REQUEST:',r
@@ -289,21 +346,35 @@ class XSSspider(CrawlSpider):
return (hostname, protocol, doc_domain, path)
def xss_reqs(self, response):
def url_xss_reqs(self, response):
''' Check for injection locations in app '''
orig_url = response.meta['resp_url']
payload = response.meta['payload']
body = response.body
doc = lxml.html.fromstring(body)
url = response.url
payload = response.meta['payload']
orig_url = response.meta['orig_url']
quote_enclosure = self.single_or_double_quote(body)
injections = self.inj_points(payload, doc) # XSS can only occur within HTML tags or within an attribute
payloads = self.xss_str_generator(injections, quote_enclosure)
for i in injections:
print i
for p in payloads:
print p
urls = [(orig_url.replace(self.test_str, p), p) for p in payloads]
for u in urls:
print u
# dont_filter is necessary since scrapy sees urlencoded payloads as the same as unencoded
reqs = [Request(url[0], callback=self.xss_chars_finder, meta={'payload':url[1], 'injections':injections, 'quote':quote_enclosure}, dont_filter=True) for url in urls]
reqs = [Request(url[0],
callback=self.xss_chars_finder,
meta={'payload':url[1],
'injections':injections,
'quote':quote_enclosure,
'type':'url'},
dont_filter=True)
for url in urls]
return reqs
@@ -324,19 +395,25 @@ class XSSspider(CrawlSpider):
return sorted(injections)
def xss_str_generator(self, injections, quote_enclosure):
''' This is where the injection points are analyzed and specific payloads are created '''
attr_pld = quote_enclosure+self.tag_pld
payloads = []
for i in injections:
print 'POSTinj'
print i ##################3
line = i[0]
tag = i[1]
# Attribute XSS payloads
if len(i) > 2:
attr = i[2]
attr_val = i[3]
# Test for open redirect/XSS
if self.redir_pld not in payloads:
if attr == 'href' and attr_val == self.test_str:
if self.redir_pld not in payloads:
payloads.append(self.redir_pld)
# Test for normal attribute-based XSS (needs either ' or " to be unescaped depending on which char the value is wrapped in
@@ -345,11 +422,13 @@ class XSSspider(CrawlSpider):
payloads.append(attr_pld)
continue
if self.tag_pld not in payloads:
# Between tag XSS payloads
else:
# Test for embedded js xss
if tag == 'script' and self.js_pld not in payloads:
payloads.append(self.js_pld)
# Test for normal between tag XSS (no quotes necessary)
if self.tag_pld not in payloads:
payloads.append(self.tag_pld)
if self.tag_pld in payloads and attr_pld in payloads:
@@ -374,25 +453,21 @@ class XSSspider(CrawlSpider):
found_chars = set()
body = response.body
resp_url = response.url
try:
xss_type = response.meta['type']
if xss_type == 'url':
url = re.sub(self.test_str+'.*'+self.test_str, 'INJECT', resp_url)
except Exception:
else:
url = resp_url
injections = response.meta['injections']
quote_enclosure = response.meta['quote']
chars_between_delims = '%s(.*?)%s' % (self.test_str, self.test_str)
payload = response.meta['payload'].strip(self.test_str) # xss char payload
if '%' in payload:
payload = urllib.unquote_plus(payload)
quote_enclosure = response.meta['quote']
chars_between_delims = '%s(.*?)%s' % (self.test_str, self.test_str)
inj_num = len(injections)
xss_num = 0
# Usually embedded JS will use the opposite quote as html attribute quotes
#if quote_enclosure == '"':
# jsq = "'"
#else:
# jsq = '"'
break_tag_chars = set(['>', '<',])
break_attr_chars = set([quote_enclosure])
break_js_chars = set(['"', "'", ';'])
@@ -400,6 +475,7 @@ class XSSspider(CrawlSpider):
# Check the entire body for exact match
if payload in body:
msg = '%s | No filtered XSS characters | Unfiltered: %s' % (url, payload)
item['type'] = xss_type
item['vuln_url'] = msg
return item
@@ -408,6 +484,7 @@ class XSSspider(CrawlSpider):
xss_num = len(matches)
if xss_num > 0:
if xss_num != inj_num:
item['type'] = xss_type
err = ('%s | Mismatch between harmless injection count and payloaded injection count: %d vs %d' % (url, inj_num, xss_num))
item['error'] = err
@@ -420,34 +497,40 @@ class XSSspider(CrawlSpider):
found_chars = set()
for i in injections:
attr = None
attr_val = None
line = i[0]
tag = i[1]
if len(i) > 2: # attribute injections have 4 data points within this var
attr = i[2]
attr_val = i[3]
for c in all_found_chars:
for c in all_found_chars: # c = set of characters found at injection point
chars = set(c)
joined_chars = ''.join(chars)
if quote_enclosure in payload: # attr
if quote_enclosure in payload and attr: # attr
if break_attr_chars.issubset(chars):
msg = '%s | Able to break out of attribute | Unfiltered: %s' % (url, joined_chars)
item['type'] = xss_type
item['vuln_url'] = msg
if '<' and '>' in payload: # tag
if '<' and '>' in payload and not attr: # tag
if break_tag_chars.issubset(chars):
msg = '%s | Able to break out of tags | Unfiltered: %s' % (url, joined_chars)
item['type'] = xss_type
item['vuln_url'] = msg
if ';' in payload: #js
if break_js_chars.issubset(chars):
msg = '%s | Able to break out of javascript | Unfiltered: %s' % (url, joined_chars)
item['type'] = xss_type
item['vuln_url'] = msg
if 'javascript:prompt(99)' == payload.lower():
if 'javascript:prompt(99)' == payload.lower(): #redir
if 'javascript:prompt(99)' == joined_chars.lower():
msg = '%s | Open redirect/XSS vulnerability | Unfiltered: %s' % (url, joined_chars)
item['type'] = xss_type
item['vuln_url'] = msg
def parse_attr_xpath(self, xpath):
@@ -510,16 +593,3 @@ class XSSspider(CrawlSpider):
# If there's multiple injection points of the same kind, make a list of them and return the highest value option
# match the injection point type with the injection payload type, run through all the matches putting the results into a list, then scan the list
# for the highest danger injection point + payload and return that as item['level']
#def dupe_form(self, values, url, method):
# ''' True if the form values/url/method are identical to previously sent form '''
# values = tuple(values)
# vam = (values, url, method)
# vamset = set(vam)
# print '---VAM', vam
# if vamset.issubset(self.form_requests_made):
# print '----------------------- ALREADY SENT THIS ONE ----------------------'
# return True
# self.form_requests_made.add(vam)
# return