Home

Awesome

MITM-cheatsheet

We tried to put together all known MITM attacks and methods of protection against these attacks. Here is also contains tools for carrying out MITM attacks, some interesting attack cases and some tricks associated with them.

Note: Almost all attack tools, described here, doesn't have any sniffer inside. It only provides attack. Tools for sniffing here: Data sniffing.

A cheat sheet for pentesters and defensive teams about Man In The Middle attacks.

Table of Contents

:hourglass_flowing_sand: - part in process

L2

Arp spoofing

Сomplexity: Low
Relevance: High
Description:
Address Resolution Protocol (ARP) designed for resolving IP addresses to MAC addresses. All network devices that need to communicate in the network use broadcast ARP queries to find out other machines’ MAC addresses.

Almost all arp spoofing tools use a gratuitous arp response. A Gratuitous ARP reply is a reply to without a ARP request.

Despite the effectiveness of gratuitous ARP, it is particularly insecure because it can be used to assure the remote host that the MAC address of a system on the same network has changed and to specify which address is used now.

<details> <summary>The typical example of arp spoofing attack</summary>

  1. Before ARP-spoofing is performed, there are entries in the ARP tables of nodes A and B with IP and MAC addresses of each other. The information is transmitted between nodes A and B.

  2. During the ARP-spoofing process, the С computer performing the attack sends ARP responses (without receiving requests = gratuitous arp) =>

  1. As the computers support gratuitous ARP, they modify their own ARP tables and place records where the MAC address of computer C is instead of the real MAC address of computer A and B.
</details>

Also there is a chance of successful attack in another way. When you will monitor the arp activity in the network segment and suddenly notice the victim's arp request, you can try send the arp reply to victim faster than addressee of that request. Some vendors can accept this trick.

Attack tools:

Attack detection

Attack prevention

Related links
How ARP works

STP(RSTP, PVSTP, MSTP) spoofing

Сomplexity: High
Relevance: Moderate
Description:
The spanning tree protocol is designed to detect and prevent loops in the network if there are redundant paths between the switches.

Anyone who can emulate a device with a (lower) root switch identifier (by connecting a new virtual device with a lower priority or using the STP packet generation tool) can partially or completely intercept the virtual network traffic. Typically, an attacker does not have a physical connection with two switches, so the described attack method is hardly possible. However, in wireless networks, the situation is changing, since the cable connection (socket in the office) and the wireless connection (access point) can end on different switches.

Attack tools
Caution: Often this type of attack leads to a denial of service.

Defence technics

NDP spoofing

Сomplexity: Moderate
Relevance: Close to None
Description:
Attack tools
Defence technics

VLAN hopping

Сomplexity: Moderate
Relevance: None
Description:
A virtual LAN (Local Area Network) is a logical subnetwork that can group together a collection of devices from different physical LANs. Larger business computer networks often set up VLANs to re-partition their network for improved traffic management.

VLANs work by applying tags to network frames and handling these tags in networking systems – creating the appearance and functionality of network traffic that is physically on a single network but acts as if it is split between separate networks.

VLAN hopping is a common name for attacks that involve access to the VLAN, which was initially (before the attack) unavailable to the attacker.

<details> <summary>It could be performed in two ways: </summary>

  1. The primary VLAN Hopping attack (using DTP)
    Works only on old Cisco switches.
    An attacker acts as a switch in order to trick a legitimate switch into creating a trunking link between them. Packets from any VLAN are allowed to pass through a trunking link. Once the trunk link is established, the attacker then has access to traffic from any VLAN. This method is only successful when the legitimate switch is configured to negotiate a trunk. This occurs when an interface is configured with either "dynamic desirable", "dynamic auto" or "trunk" mode. If the target switch has one of those modes configured, the attacker then can generate a DTP message from their computer and a trunk link can be formed.

  2. Double tagging occurs when an attacker adds and modifies tags on an Ethernet frame to allow the sending of packets through any VLAN. This attack takes advantage of how many switches process tags. Most switches will only remove the outer tag and forward the frame to all native VLAN ports. With that said, this method is only successful if the attacker belongs to the native VLAN of the trunk link. Another important point is, this attack is strictly one way as it is impossible to encapsulate the return packet. </details>

The Exploit-db doc
The Guide with illustrations and video
VLAN hopping full guide
In-depth article

Attack tools

  1. The primary VLAN Hopping attack (using DTP)
    It can only be performed when interfaces are set to negotiate a trunk. To prevent the VLAN hopping from being exploited, we can do the below mitigations:

    • Ensure that ports are not set to negotiate trunks automatically by disabling DTP
    • Do not configure any access points with either of the following modes: "dynamic desirable", "dynamic auto", or "trunk".
    • Shutdown all interfaces that are not currently in use.

  2. Double Tagging
    To prevent a Double Tagging attack, keep the native VLAN of all trunk ports different from user VLANs.

L3

SLAAC Attack

Complexity: Low
Relevance: High
Description
SLAAC - Stateless Address AutoConfiguration. SLAAC os the one of ways of host network configuration, like DHCPv4. SLAAC provides an IPv6 host prefix value, prefix length and default gateway link-local address without DHCPv6-server which keeps state of the provided addresses (thats why it's called stateless). The SLAAC process is performed during SLAAC-only and SLAAC+DHCPv6 Stateless configuration.

The main problem of this process is that the attacker can craft the rogue RA to give the hosts his own configuration (e.g., to become a default router on the link). All the hosts, which have IPv6 enabled, are potentially vulnerable to SLAAC attacks. Especially in cases, when IPv6 is enabled is OS by default but organization hasn't deployed IPv6 in any form.

Another threat in RA comes from the ability to send DNS configuration over RA, so that attacker can spoof it, too: RFC 6106 - IPv6 Router Advertisement Options for DNS Configuration.

Attack Tools

Defence technics
The simpliest way to mitigate SLAAC-attacks is to just disable IPv6 on all hosts in the network. But this solution is only suitable for networks where IPv6 stack is not in use and was enabled just due to misconfiguration.

<details> <summary>Vendors implementation</summary>

Cisco has implemented a technology "IPv6 First Hop Security" which is included in Catalyst 6500, 4500, 3850, 3750 and 2960 Series Switches, 7600 Series Routers and Cisco 5700 Series Wireless LAN Controllers. There's RA Guard, DHCP Guard and also IPv6 Snooping implemented. More information can be found here.

Juniper has implemented RA Guard. There is one strange fact: on the router-advertisement-guard statement documentation page it's mentioned that only EX Series platforms are supported. But on the page of Configuring Stateless IPv6 Router Advertisement Guard and Configuring Stateful IPv6 Router Advertisement Guard it's mentioned that both EX and some of QFX Series platforms support RA Guard: EX2300(15.1X53-D56), EX2300-VC(15.1X53-D56), EX3400(15.1X53-D55), EX3400-VC(15.1X53-D55), EX4300(16.1R1), EX4300-VC(16.1R1), EX4300 Multigigabit(18.2R1), EX4600(18.3R1), EX4600-VC(18.3R1) and QFX5100(18.2R1), QFX5110(17.2R1), QFX5200(18.2R1).

Mikrotik, unfortunately, hasn't implemented such technologies. Theres's a presentation from Mikrotik Users' Meeting and author advised just to isolate Layer 2 segment of network. No other valuable advices were found. The problem was also mentioned on the Mikrotik users' forum in 2012.

Unfortunately, there are methods of traffic analysis hardening, which breaks performance of protections techniques (e.g. hiding the RA in Hob-By-Hop header). There is a draft RFC which describes the evasion of RA Guard. The evasion technique is based on usage of IPv6 packet fragmentation. Some additional recommendations on fragmentation are presented in RFC 6980 - Security Implications of IPv6 Fragmentation with IPv6 Neighbor Discovery.

</details> <details> <summary>10 basic ideas to solve the problem</summary>

RFC 6104 - Rogue IPv6 Router Advertisement Problem Statement presents 10 basic ideas to solve the problem of Rogue RA. So the section above is just a brief overview of what IETF has to offer as a solution for today:

  1. Manual Configuration of IPv6 address and disabling autoconfiguration for RA messages to be ignored.
    For Linux systems net.ipv6.conf.* values can be changed:

    net.ipv6.conf.all.autoconf = 0  
net.ipv6.conf.all.accept_ra = 0  
net.ipv6.conf.default.accept_ra=0  
net.ipv6.conf.all.accept_ra=0  
net.ipv6.conf.eth0.accept_ra=0

    For Mac-OS there is a guide for IPv6 hardening. But the author faced a problem with parameter responsible for acceptance of RAs in Mac-OS: net.inet6.ip6.accept_rtadv must be set to 0 but its impossible. It's called deprecated in kernel source code and is defined as read-only, but Mac-OS keeps accepting RAs. So, in Mac-OS it's not possible to disable RAs through sysctl. The one thing that can be done is setting up the maximum number of acceptable prefixes and maximum number of acceptable default routers to 1.

    For Windows there is a command which can be run under admininstrator to disable autoconfoguration:

    netsh interface ipv6 set interface "Local Area Connection" routerdiscovery=disabled

  2. RA Snooping in L2 switches similarly to DHCP snooping, so that RAs from wrong sources can be dropped.

  3. ACLs on Managed Switches can be used if there is a mechanism of ACL on a switch which can block ICMPv6 RA outbound on user ports(used to access LAN by users). So if such ACL is possible to implement on a used platform, no user on LAN will be able to broadcast/unicast RA.

  4. SEcure Neighbor Discovery - SEND - RFC 3971 is a protocol, which offers the use of public key cryptography to secure the communications between router and hosts.

  5. Router Preference Option - this method is only suitable in case of accidental RAs from users. The idea is that administrator can set "High" level of preference in all legitimate RAs so that IPv6 hosts wont overwrite the configuration received by such RAs if they have "Medium" or "Low" preference level. The Router Preference Option is present in RFC 4191 - Default Router Preferences and More-Specific Routes.

  6. Rely on Layer 2 Admission Control - the idea is based on relying on deployment of 802.1x so that attackers won't be able to join LAN to send RAs and perform attack.

  7. Using Host-Based Packet Filters - if there is an ability to push configuration to users' machines, the host-based packet filters can be configured to accept RAs only from exact IPv6 addresses.

  8. Using an "Intelligent" Deprecation Tool - the idea is to observe the link traffic for rogue RAs and to deprecate them for hosts by sending a deprecating RA with rogue router's address in it and router lifetime field set to 0. Attack the attack's traffic.

  9. Using Layer 2 Partitioning - the idea is that if each user or system is partitioned into a different Layer 2 medium the impact if some rogue RA can be limited. This method causes software and hardware costs growing.

  10. Adding Default Gateway/Prefix Options to DHCPv6 - leaving SLAAC autoconfiguration for DHCPv6 autoconfiguration partly solves the problem of default gateways and prefixes sent by rogue RAs but also leads to problems with rogue DHCPv6 servers. The second problem is that RA is still used to inform hosts to use DHCPv6.

The 4th section of RFC 6104 has a table which contains the ways of mitigation suitability for 2 cases of Rogue RA: administrator's mistake and user's mistake.

</details>

Related Monitoring Tools
There are some tools, which can be helpful in rogue RA detection and monitoring:

Related RFCs
RFC 6104 - Rogue IPv6 Router Advertisement Problem Statement
RFC 6105 - IPv6 Router Advertisement Guard
RFC 3736 - Stateless Dynamic Host Configuration Protocol (DHCP) Service for IPv6
RFC 4862 - IPv6 Stateless Address Autoconfiguration (SLAAC)
RFC 7113 - Implementation Advice for IPv6 Router Advertisement Guard (RA-Guard)
RFC 8021 - Generation of IPv6 Atomic Fragments Considered Harmful

Other useful related links
Windows machines compromised by default configuration flaw in IPv6
Why You Must Use ICMPv6 Router Advertisements

Hijacking HSRP (VRRP, CARP)

Сomplexity: High
Relevance: High
Description:
Attack tools Scapy is the easiest way to create PoC and hijack active node status:

For HSRP:

#!/usr/bin/env python
from scapy.all import *

if __name__ == "__main__":
    ip = IP(src="10.0.0.100", dst="224.0.0.2")
    udp = UDP(sport=1985, dport=1985)
    hsrp = HSRP(group=1, priority=150, virtualIP="10.0.0.1", auth="cisco")
    send(ip/udp/hsrp, iface="eth1", inter=3, loop=1)

For VRRP:

#!/usr/bin/env python
from scapy.all import *

if __name__ == "__main__":
    ip = IP(src="10.0.0.100", dst="224.0.0.2")
    udp = UDP()
    vrrp = VRRP(vrid=1, priority=150, addrlist=["10.0.0.7", "10.0.0.8"], ipcount=2, auth1='cisco')
    send(ip/udp/vrrp, iface="eth1", inter=3, loop=1)

Defence technics

Dynamic routing protocol spoofing (BGP)

Сomplexity: High
Relevance: High
Conditions:
Description:
Attack tools
https://github.com/fredericopissarra/t50
Defence technics

RIPv2 Routing Table Poisoning

Сomplexity: Medium
Relevance: Medium
Conditions:
RIP implemented;
RIPv1 in use;
RIPv2 authentication disabled.
Description:
There are 3 versions of RIP:

The most widely implemented protocol is RIPv2. RIPv1 is not secure at all, as it doesn't support message authentication. There is a good write up on exploiting RIPv1 by injecting a fake route.

As specified in RFC 2453, RIPv2 router must exchange routing information every 30 seconds. The idea of attack is to send fake RIP Response messages, which contain the route an attacker needs to inject. Though, there is a special multicast for RIPv2 routers - 224.0.0.9, responses, sent as unicast can be accepted, too. This, for example may harden the detection of the attack, comparing to the case of multicast fake routing propagation. There is a good short write up on exploting RIPv2 network with no RIPv2 authentication with Scapy usage example.

Attack tools

Defence technics
If router is not configured to authenticate RIPv2 messages, it will accept RIPv1 and RIPv2 unauthenticated messages. The most secure configuration in this way is to set up RIPv2 authentication so that router should not be accepting RIPv1 and v2 unauthenticated messages and so making an unauthenticated router unable to inject a route. This mechanism is described in RFC 2082 - RIP-2 MD5 Authentication, but, it describes the usage of MD5, which is acknowledged to be a weak hashing function. The better one, which means the use of SHA-1 is described in RFC 4822 - RIPv2 Cryptographic Authentication.

Unfortunately, RIPv2 supports only Plain-text and MD5 Authentication. The first one is useless in case of sniffing the network, MD5 Auth is better in case of a passive attacker, intercepting packets, as it doesn't transfer password in plain text.
The Configuration of RIPv2 Authentication guide describes how to set this feature on Cisco devices.
The guide for setting MD5 authentication for Mikrotik is present here. The guide for setting MD5 authentification on Juniper devices is present here.

Also, passive-interface feature should be used on the access interfaces, which communicate to end devices.
Mikrotik's documentation on setting passive interface feature.
Cisco's documentation on setting passive interface feature.
Juniper's documentation on setting passive-interface feature.

Related RFCs:
RFC 1388 - RIP Version 2 Carrying Additional Information
RFC 4822 - RIPv2 Cryptographic Authentication
RFC 2453 - RIP Version 2
RFC 2080 - RIPng for IPv6.

OSPF Routing Table Poisoning

Сomplexity: High
Relevance: High
Conditions:
Description:
Attack tools
Defence technics

EIGRP Routing Table Poisoning

Complexity: Medium
Relevance: Medium
Conditions: EIGRP protocol implemented on the network; no EIGRP messages authentication set up
Description:
EIGRP stands for Enhanced Interior Gateway Routing Protocol. It is a proprietary Cisco’s distance vector routing protocol, relying on Diffused Update Algorithm - DUAL. The main purpose of this protocol is to dynamically update the routing table and propagate the routes to other routers.
The main security issue is possible in case of spoofing data in Update message, e.g. to inject a non-legitimate route. In this case the router's routing table gets changed to make it pass the traffic through the device, controlled by the attacker and so the MitM attack is present.

Attack tools

Defence techniques
To protect a network from untrusted route propagations, EIGRP provides a mechanism for authenticating router updates. It uses MD5-keyed digest to sign each packet to prevent unauthorized devices from sending updates to the network. It protects legitimate routers from non-legitimate router updates and from router spoofing. The key is just a defined string, which must be set on other devices which are meant to be legitimate. The detailed guide on EIGRP MD5 Authentication setup can be found here.

Unfortunately, MD5 is acknowledged to be a weak hashing algorithm due to hash collisions. Cisco devices also support hmac-sha-256 EIGRP Updates Authentification. The hash collision attack on SHA-256 is much more complex than for MD5. The guide to EIGRP HMAC-SHA-256 Authentication can be found here.

The stub EIGRP routing area can be set up as it let's determine the types of routes the stub router should receive queries or not. More information on EIGRP Stub Routing can be found here.

Another best practice to reduce unwanted traffic in a network is to set up passive interfaces. passive-interface feature should be set on access interfaces, which communicate not to network devices, but to end devices. The instruction on setting passive-interface on EIGRP and explaination on how it works is presented in Cisco's documentation page.

ICMP Redirect

Сomplexity: Medium
Relevance: Medium
Description:
One of the purposes of the ICMP Protocol is to dynamically change the routing table of the end network systems. Dynamic remote management routing was originally conceived to prevent possible send a message to a non-optimal route, as well as to increase fault tolerance of the Network as a whole. It was assumed that the network segment can be connected to the Internet through several routers (not through one as it usually happens). In this case, we can address the external network through any of the nearest routers. For example, to some_host.site the shortest route passes through the "router A" and to the another.site - through the "router B".

If one of the routers fails, communication with the outside world is possible through another router. As the "ICMP Redirest attack", we change the route to some site (DNS Name) in the routing table of node A (victim) so that the traffic from node A to some site goes through hacker PC

Conditions of success:

Attack tools

Defence technics

L4+

NetBIOS (LLMNR) spoofing

Сomplexity: Low
Relevance: High
Description:
If a windows client cannot resolve a hostname using DNS, it will use the Link-Local Multicast Name Resolution (LLMNR) protocol to ask neighbouring computers. LLMNR can be used to resolve both IPv4 and IPv6 addresses.

If this fails, NetBios Name Service (NBNS) will be used. NBNS is a similar protocol to LLMNR that serves the same purpose. The main difference between the two is NBNS works over IPv4 only.

The problem of this pretty cool thing is that when LLMNR or NBNS are used to resolve a request, any host on the network who knows the IP of the host being asked about can reply. Even if a host replies to one of these requests with incorrect information, it will still be regarded as legitimate.

The attacker may request NTLM authentication from the victim, which will cause the victim's device to send an NTLM hash, which can then be used for brute force attack.

<details> <summary>Also there is a chance to perform WPAD spoofing.</summary>

WPAD spoofing can be referred to as a special case of LLMNR- and NBNS-spoofing. Web Proxy Auto Discovery protocol is used for automatic configuration of HTTP proxy server.

The device sends an LLMNR/NBNS request with a wpad host, obtains the corresponding IP address and tries to access the wpad.dat file containing information about proxy settings via HTTP.

As a result, an attacker can perform LLMNR/NBNS spoofing and provide the victim with his own wpad.dat file, resulting in all HTTP and HTTPS traffic going through the attacker.</details>

Quick tutorial to grab clear text credentials
How Microsoft Windows’s name resolution services work and how they can be abused

Attack tools

Defence technics

DHCP spoofing

Сomplexity: Moderate
Relevance: Moderate
Description:
The purpose of this attack is to use the attacker's host or device as the default gateway and to force clients to use a false Domain Name Service (DNS) and a Windows Internet name service (WINS server) configured by the attacker. The attacker's task is to configure a fake DHCP server on the network to provide DHCP addresses to clients and exhausted the pool of IP addresses from other legitimate DHCP servers (DHCP Starvation attack).

Conditions of success:

DHCP Starvation attack:

Attack tools for DHCP starvation

Attack tools for DHCP spoofing

Defence technics

This is a L2 switch function designed to protect against DHCP attacks. For example, a DHCP spoofing attack or DHCP starvation attack.

On Cisco Switches:

Important. By default, after enabling DHCP snooping, the switch is enabled to check for MAC address matching. The switch checks whether the MAC address in the DHCP request matches the client's MAC address. If they do not match, the switch discards the packet.

How DHCP works
DHCP wireshark sample

Rogue DHCP (DHCPv6)

Сomplexity: Low
Relevance: High
Description:
The Ipv6 client sends a Solicit message to the All_DHCP_Relay_Agents_and_Servers address to find available DHCP servers. Any server that can meet the client's requirements responds with an Advertise message. The client then chooses one of the servers and sends a Request message to the server asking for confirmed assignment of addresses and other configuration information.The server responds with a Reply message that contains the confirmed addresses and configuration.

This schema looks simular to DHCPv4 so the main goal for the attacker is to use fake DHCPv6 server to redirect victims traffic to himself.

The attacker can catch client DHCP solicit message and can actually reply, pretending that he is the DHCPv6 server and assign credentials (such as the DNS address) to be used by victim.

Attack tools

Defence technics

Wireless

Karma attacks (Wi-Fi)

Сomplexity: Low
Relevance: High
Description:
The KARMA attack uses the peculiarities of the clients who send requests to determine which wireless networks are nearby.

The Wi-Fi Access Point periodically sends a beacon request indicating the network SSID that identifies the Wi-Fi network. When a client receives a beacon frame with an SSID that it remembers, it can be associated with the wireless network. Vulnerable client devices broadcast a "preferred network list" (PNL), which contains the SSIDs of access points to which they have previously connected and are willing to automatically reconnect without user intervention. These broadcasts may be received by any WiFi access point in range. The KARMA attack consists in an access point receiving this list and then giving itself an SSID from the PNL, thus becoming an evil twin of an access point already trusted by the client.

As a result, the client connects to a different network than the one the user expects. And now the attacker can perform MITM or other attacks on the client system.

However, nowadays, most modern network managers have taken countermeasures against the KARMA attack by switching to passive scanning; instead of arbitrarily sending probe request frames, network managers now wait to receive a beacon frame with a familiar ESSID before associating with a wireless network. While this countermeasure has hampered the effectiveness of the KARMA attack, the second feature exploited by KARMA, the Auto-Connect flag that enables the stations to automatically join previously connected networks, was left intact in almost every modern Operating System.

An attacker that can guess the SSID in the victim device's Preferred Network List, will be able to broadcast the corresponding beacon frame and have that device automatically associate with an attacker-controlled access point. In a more sophisticated version of the attack, the adversary may use a "dictionary" of common SSIDs, that the victim has likely connected to in the past.

How does a KARMA attack work?

Attack tools

Defence technics

Rogue BTS (GSM)

Attack technics

Data sniffing

Attack tools:

Injections in data

Malicious JS in HTML

HTA

Data modification

Wsus

DNS hijacking

Hacker notes

Difference between technology stack (or why most of that attack imposible from your notebook)

Thanks to @serjepatoff for explanation:
Network equipment has the same general-purpose CPUs. MIPS or ARM for basic SOHO routers, multicore x86 in more serious boxes. It's the NIC with multiple hardware queues and special methods of NIC<->software communication (explicit ring–buffering, heavy use of DMA) that make a difference.

So you can't just connect to the network and turn on the spoofing, it can put the network down right away. Your small notebook network adapter simply cannot cope with a large data stream and will start to drop them. You need to choose the optimal number of hosts that you can spoof at the same time(~<4).

Attack device

Possible candidate: MikroTik hAP AC

SSLStrip, SSLStrip+, HSTS

SSLStrip is a technique that replaces a secure (HTTPS) connection with an open (HTTP) connection. This attack is also known as HTTP-downgrading

It intercepted HTTP traffic and whenever it spotted redirects or links to sites using HTTPS, it would transparently strip them away.

Instead of the victim connecting directly to a website; the victim connected to the attacker, and the attacker initiated the connection back to the website. The interceptor made the encrypted connection to back to the web server in HTTPS, and served the traffic back to the site visitor unencrypted

But it doesn't work anymore with the advent of HSTS. More precisely, it doesn't work where HSTS is enabled.

HTTP Strict Transport Security (HSTS) is a web security policy mechanism that helps to protect websites against protocol downgrade attacks (SSL stripping). It allows web servers to declare that web browsers should interact with it using only secure HTTPS connections, and never via the insecure HTTP protocol. HSTS is an IETF standards track protocol and is specified in RFC 6797.

The HSTS works by the server responding with a special header called Strict-Transport-Security which contains a response telling the client that whenever they reconnect to the site, they must use HTTPS. This response contains a "max-age" field which defines how long this rule should last for since it was last seen.

Also It has includeSubDomains (optional). If this optional parameter is specified, this rule applies to all of the site's subdomains as well.

But not everyone sets up HSTS the same way.

That's how SSLstrip++ came about.

It's a tool that transparently hijack HTTP traffic on a network, watch for HTTPS links and redirects, then map those links into either look-alike HTTP links or homograph-similar HTTPS links.

One of the shortcomings of HSTS is the fact that it requires a previous connection to know to always connect securely to a particular site. When the visitor first connects to the website, they won't have received the HSTS rule that tells them to always use HTTPS. Only on subsequent connections will the visitor's browser be aware of the HSTS rule that requires them to connect over HTTPS.

HSTS Preload Lists are one potential solution to help with these issues, they effectively work by hardcoding a list of websites that need to be connected to using HTTPS-only.

Inside the source code of Google Chrome, there is a file which contains a hardcoded file listing the HSTS properties for all domains in the Preload List. Each entry is formatted in JSON.

Attack tools

Defence technics

For developers - The 6-Step "Happy Path" to HTTPS