#!/usr/bin/ruby # # This script leverages couple of methods in order to validate that passed # domain is a S3 bucket indeed. # # Mariusz B., 2019, # require 'resolv' require 'uri' require 'net/http' DEBUG = false $cached_responses = {} $dns_records = {} $random_resource = (0...32).map { ('a'..'z').to_a[rand(26)] }.join class Resp attr_accessor :body attr_accessor :headers def to_s return @body end def to_str return @body end end def dbg(x) if DEBUG puts "[dbg] #{x}" end end def checkDnsRecords(bucket) begin Resolv::DNS.open do |dns| $dns_records['ip'] = dns.getaddress(bucket).to_s $dns_records['rev-dns'] = dns.getnames($dns_records['ip']).pop.to_s end rescue Resolv::ResolvError dbg "\tCould not resolve name #{bucket}." return false end if $dns_records['rev-dns'].end_with? '.amazonaws.com' and $dns_records['rev-dns'].include? 's3' dbg "\tReverse-DNS record for IP (#{$dns_records['ip']}) points to AWS S3: #{$dns_records['rev-dns']}" return true end return false end def fetch(url) unless $cached_responses.key? url begin uri = URI.parse(url) response = Net::HTTP.get_response uri resp = Resp.new resp.body = response.body resp.headers = response.each_header.to_h $cached_responses[url] = resp rescue Exception => e #puts "\tHTTP Request (#{url}) failed: #{e}" $cached_responses[url] = nil end end return $cached_responses[url] end def checkServerHeader(bucket) ['http', 'https'].each do |scheme| out = fetch "#{scheme}://#{bucket}" if not out.nil? and out.headers.include? 'server' and out.headers['server'].downcase == 'amazons3' dbg "\tAmazon S3 bucket found by 'Server' HTTP response header contents." return true end end ['http', 'https'].each do |scheme| out = fetch "#{scheme}://#{bucket}.s3.amazonaws.com" if not out.nil? and out.headers.include? 'server' and out.headers['server'].downcase == 'amazons3' dbg "\tAmazon S3 bucket found by 'Server' HTTP response header contents." return true end end ['http', 'https'].each do |scheme| out = fetch "#{scheme}://s3.amazonaws.com/#{bucket}" if not out.nil? and out.headers.include? 'server' and out.headers['server'].downcase == 'amazons3' dbg "\tAmazon S3 bucket found by 'Server' HTTP response header contents." return true end end return false end def checkAmzHeaders(bucket) out = fetch "http://#{bucket}.s3.amazonaws.com" if not out.nil? and out.headers.include? 'x-amz-request-id' and out.headers.include? 'x-amz-id-2' dbg "\tAmazon S3 found by 'x-amz-request-id' and 'x-amz-id-2' HTTP response headers existence." return true end out = fetch "http://s3.amazonaws.com/#{bucket}" if not out.nil? and out.headers.include? 'x-amz-request-id' and out.headers.include? 'x-amz-id-2' dbg "\tAmazon S3 found by 'x-amz-request-id' and 'x-amz-id-2' HTTP response headers existence." return true end return false end def checkBucketRegionHeader(bucket) out = fetch "http://#{bucket}.s3.amazonaws.com" if not out.nil? and out.headers.include? 'x-amz-bucket-region' dbg "\tAmazon S3 bucket region found in 'x-amz-bucket-region' HTTP response header" return true end out = fetch "http://s3.amazonaws.com/#{bucket}" if not out.nil? and out.headers.include? 'x-amz-bucket-region' dbg "\tAmazon S3 bucket region found in 'x-amz-bucket-region' HTTP response header" return true end return false end def checkBucketResponse(bucket) traces = [ 'ListBucketResult xmlns=', '', 'AccessDeniedAccess Denied', 'AllAccessDisabled', 'PermanentRedirect', '', 'TemporaryRedirect', "#{bucket}", "#{bucket}" ] out = fetch "http://#{bucket}.s3.amazonaws.com" if not out.nil? and out.headers.include? 'content-type' and out.headers['content-type'].downcase == 'application/xml' traces.each do |trace| if out.body.include? trace dbg "\tAmazon S3 bucket identified by trace in body: '#{trace}'" return true end end end out = fetch "http://s3.amazonaws.com/#{bucket}" if not out.nil? and out.headers.include? 'content-type' and out.headers['content-type'].downcase == 'application/xml' traces.each do |trace| if out.body.include? trace dbg "\tAmazon S3 bucket identified by trace in body: '#{trace}'" return true end end end return false end def checkNonExistentResourceBucketResponse(bucket) traces = [ '
  • Code: NoSuchKey
  • ', 'NoSuchKey', ] out = fetch "http://#{bucket}.s3.amazonaws.com/#{$random_resource}" unless out.nil? traces.each do |trace| if out.body.include? trace dbg "\tAmazon S3 bucket identified by trace in body of a non-existent resource: '#{trace}'" return true end end end out = fetch "http://s3.amazonaws.com/#{bucket}/#{$random_resource}" unless out.nil? traces.each do |trace| if out.body.include? trace dbg "\tAmazon S3 bucket identified by trace in body of a non-existent resource: '#{trace}'" return true end end end return false end def checkIfBucketExists(bucket) traces = [ 'NoSuchBucket', 'The specified bucket does not exist', 'flaws.cloudfsdsdfsdfdsf' ] found = 0 out = fetch "http://#{bucket}.s3.amazonaws.com" if not out.nil? and out.headers.include? 'content-type' and out.headers['content-type'].downcase == 'application/xml' traces.each do |trace| if out.body.include? trace found += 1 end end end if found == traces.length dbg("Bucket verified to be non-existent.") return false end out = fetch "http://s3.amazonaws.com/#{bucket}" if not out.nil? and out.headers.include? 'content-type' and out.headers['content-type'].downcase == 'application/xml' traces.each do |trace| if out.body.include? trace found += 1 end end end if found == traces.length dbg("Bucket verified to be non-existent.") return false end return true end def main(args) puts %{ :: Identifies AWS S3 Buckets via couple of methods Mariusz B. 19', } if ARGV.length != 1 puts "Usage: ./identifyS3Bucket.rb " exit end points = 0 bucket = ARGV.pop puts "[.] Examining bucket with name: #{bucket}" unless checkIfBucketExists bucket puts "[-] There is no such bucket." exit 1 end if checkDnsRecords bucket puts "[+] S3 bucket identified via DNS records." points += 1 end if checkServerHeader bucket puts "[+] S3 Bucket identified by HTTP header 'Server' in response." points += 1 end if checkAmzHeaders bucket puts "[+] S3 Bucket identified by HTTP amz headers." points += 1 end if checkBucketResponse bucket puts "[+] S3 Bucket identified via traces in HTTP response body." points += 1 end if checkNonExistentResourceBucketResponse bucket puts "[+] S3 Bucket identified via traces in HTTP response of a non-existent resource." points += 1 end return 0 if points > 0 return 1 end if __FILE__ == $0 main(ARGV) end