http-csrf
NSE detection rule by George Chatzisofroniou · Same as Nmap--See https://nmap.org/book/man-legal.html
This script detects Cross Site Request Forgeries (CSRF) vulnerabilities.
It will try to detect them by checking each form if it contains an unpredictable
token for each user. Without one an attacker may forge malicious requests.
To recognize a token in a form, the script will iterate through the form's
attributes and will search for common patterns in their names. If that fails, it
will also calculate the entropy of each attribute's value. A big entropy means a
possible token.
A common use case for this script comes along with a cookie that gives access
in pages that require authentication, because that's where the privileged
exist. See the http library's documentation to set your own cookie.
description = [[
This script detects Cross Site Request Forgeries (CSRF) vulnerabilities.
It will try to detect them by checking each form if it contains an unpredictable
token for each user. Without one an attacker may forge malicious requests.
To recognize a token in a form, the script will iterate through the form's
attributes and will search for common patterns in their names. If that fails, it
will also calculate the entropy of each attribute's value. A big entropy means a
possible token.
A common use case for this script comes along with a cookie that gives access
in pages that require authentication, because that's where the privileged
exist. See the http library's documentation to set your own cookie.
]]
---
-- @usage nmap -p80 --script http-csrf.nse <target>
--
-- @args http-csrf.singlepages The pages that contain the forms to check.
-- For example, {/upload.php, /login.php}. Default: nil (crawler
-- mode on)
-- @args http-csrf.checkentropy If this is set the script will also calculate
-- the entropy of the field's value to determine if it is a token,
-- rather than just checking its name. Default: true
--
-- @output
-- PORT STATE SERVICE REASON
-- 80/tcp open http syn-ack
-- | http-csrf:
-- | Spidering limited to: maxdepth=3; maxpagecount=20; withinhost=some-very-random-page.com
-- | Found the following CSRF vulnerabilities:
-- |
-- | Path: http://www.example.com/
-- | Form id: search_bar_input
-- | Form action: /search
-- |
-- | Path: http://www.example.com/c/334/watches.html
-- | Form id: custom_price_filters
-- | Form action: /search
-- |
-- | Path: http://www.example.com/c/334/watches.html
-- | Form id: custom_price_filters
-- |_ Form action: /c/334/rologia-xeiros-watches.html
--
---
categories = {"intrusive", "exploit", "vuln"}
author = "George Chatzisofroniou"
license = "Same as Nmap--See https://nmap.org/book/man-legal.html"
local http = require "http"
local formulas = require "formulas"
local shortport = require "shortport"
local stdnse = require "stdnse"
local table = require "table"
local string = require "string"
local httpspider = require "httpspider"
portrule = shortport.port_or_service( {80, 443}, {"http", "https"}, "tcp", "open")
-- Checks if this is really a token.
isToken = function(value)
local minlength = 8
local minentropy = 72
-- If it has a reasonable length.
if #value > minlength then
local entropy = formulas.calcPwdEntropy(value)
-- Does it have a big entropy?
if entropy >= minentropy then
-- If it doesn't contain any spaces but contains at least one digit.
if not string.find(value, " ") and string.find(value, "%d") then
return 1
end
end
end
return 0
end
action = function(host, port)
local singlepages = stdnse.get_script_args("http-csrf.singlepages")
local checkentropy = stdnse.get_script_args("http-csrf.checkentropy") or false
local csrfvuln = {}
local crawler = httpspider.Crawler:new( host, port, '/', { scriptname = SCRIPT_NAME, withinhost = 1 } )
if (not(crawler)) then
return
end
crawler:set_timeout(10000)
local index, response, path
while (true) do
if singlepages then
local k, target,
k, target = next(singlepages, index)
if (k == nil) then
break
end
response = http.get(host, port, target)
path = target
else
local status, r = crawler:crawl()
-- if the crawler fails it can be due to a number of different reasons
-- most of them are "legitimate" and should not be reason to abort
if (not(status)) then
if (r.err) then
return stdnse.format_output(false, r.reason)
else
break
end
end
response = r.response
path = tostring(r.url)
end
if response.body then
local forms = http.grab_forms(response.body)
for i, form in ipairs(forms) do
form = http.parse_form(form)
local resistant = false
if form and form.action then
for _, field in ipairs(form['fields']) do
-- First we check the field's name.
if field['value'] then
resistant = string.find(field['name'], "[Tt][Oo][Kk][Ee][Nn]") or string.find(field['name'], "[cC][sS][Rr][Ff]")
-- Let's be sure, by calculating the entropy of the field's value.
if not resistant and checkentropy then
resistant = isToken(field['value'])
end
if resistant then
break
end
end
end
if not resistant then
-- Handle forms with no id or action attributes.
form['id'] = form['id'] or ""
form['action'] = form['action'] or "-"
local msg = "\nPath: " .. path .. "\nForm id: " .. form['id'] .. "\nForm action: " .. form['action']
table.insert(csrfvuln, { msg } )
end
end
end
if (index) then
index = index + 1
else
index = 1
end
end
end
-- If the table is empty.
if next(csrfvuln) == nil then
return "Couldn't find any CSRF vulnerabilities."
end
table.insert(csrfvuln, 1, "Found the following possible CSRF vulnerabilities: ")
csrfvuln.name = crawler:getLimitations()
return stdnse.format_output(true, csrfvuln)
end