Pwned

An easy, Ruby way to use the Pwned Passwords API.

API docs | GitHub repository

• Table of Contents
• About
• Installation
• Usage
  • Plain Ruby
    • Custom request options
      • HTTP Headers
      • HTTP Proxy
  • ActiveRecord Validator
    • I18n
    • Threshold
    • Network Error Handling
    • Custom Request Options
      • HTTP Headers
      • HTTP Proxy
  • Using Asynchronously
  • Devise
  • Rodauth
  • Command line
  • Unpwn
• How Pwned is Pi?
• Development
• Contributing
• License
• Code of Conduct

Troy Hunt's Pwned Passwords API allows you to check if a password has been found in any of the huge data breaches.

Pwned is a Ruby library to use the Pwned Passwords API's k-Anonymity model to test a password against the API without sending the entire password to the service.

The data from this API is provided by Have I been pwned?. Before using the API, please check the acceptable uses and license of the API.

Here is a blog post I wrote on how to use this gem in your Ruby applications to make your users' passwords better.

Add this line to your application's Gemfile:

gem 'pwned'

And then execute:

$ bundle

Or install it yourself as:

$ gem install pwned

There are a few ways you can use this gem:

1. Plain Ruby
2. Rails
3. Rails and Devise

To test a password against the API, instantiate a Pwned::Password object and then ask if it is pwned?.

password = Pwned::Password.new("password")
password.pwned?
#=> true
password.pwned_count
#=> 3303003

You can also check how many times the password appears in the dataset.

password = Pwned::Password.new("password")
password.pwned_count
#=> 3303003

Since you are likely using this as part of a sign-up flow, it is recommended that you rescue errors so if the service does go down, your user journey is not disturbed.

begin
  password = Pwned::Password.new("password")
  password.pwned?
rescue Pwned::Error => e
  # Ummm... don't worry about it, I guess?
end

Most of the times you only care if the password has been pwned before or not. You can use simplified accessors to check whether the password has been pwned, or how many times it was pwned:

Pwned.pwned?("password")
#=> true
Pwned.pwned_count("password")
#=> 3303003

You can set HTTP request options to be used with Net::HTTP.start when making the request to the API. These options are documented in the Net::HTTP.start documentation.

You can pass the options to the constructor:

password = Pwned::Password.new("password", read_timeout: 10)

You can also specify global defaults:

Pwned.default_request_options = { read_timeout: 10 }

The :headers option defines HTTP headers. These headers must be string keys.

password = Pwned::Password.new("password", headers: {
  'User-Agent' => 'Super fun new user agent'
})

An HTTP proxy can be set using the http_proxy or HTTP_PROXY environment variable. This is the same way that Net::HTTP handles HTTP proxies if no proxy options are given. See URI::Generic#find_proxy for full details on how Ruby detects a proxy from the environment.

# Set in the environment
ENV["http_proxy"] = "https://username:password@example.com:12345"

# Will use the above proxy
password = Pwned::Password.new("password")

You can specify a custom HTTP proxy with the :proxy option:

password = Pwned::Password.new(
  "password",
  proxy: "https://username:password@example.com:12345"
)

If you don't want to set a proxy and you don't want a proxy to be inferred from the environment, set the :ignore_env_proxy key:

password = Pwned::Password.new("password", ignore_env_proxy: true)

There is a custom validator available for your ActiveRecord models:

class User < ApplicationRecord
  validates :password, not_pwned: true
  # or
  validates :password, not_pwned: { message: "has been pwned %{count} times" }
end

You can change the error message using I18n (use %{count} to interpolate the number of times the password was seen in the data breaches):

en:
  errors:
    messages:
      not_pwned: has been pwned %{count} times
      pwned_error: might be pwned

If you are ok with the password appearing a certain number of times before you decide it is invalid, you can set a threshold. The validator will check whether the pwned_count is greater than the threshold.

class User < ApplicationRecord
  # The record is marked as valid if the password has been used once in the breached data
  validates :password, not_pwned: { threshold: 1 }
end

By default the record will be treated as valid when we cannot reach the haveibeenpwned.com servers. This can be changed with the :on_error validator parameter:

class User < ApplicationRecord
  # The record is marked as valid on network errors.
  validates :password, not_pwned: true
  validates :password, not_pwned: { on_error: :valid }

  # The record is marked as invalid on network errors
  # (error message "could not be verified against the past data breaches".)
  validates :password, not_pwned: { on_error: :invalid }

  # The record is marked as invalid on network errors with custom error.
  validates :password, not_pwned: { on_error: :invalid, error_message: "might be pwned" }

  # We will raise an error on network errors.
  # This means that `record.valid?` will raise `Pwned::Error`.
  # Not recommended to use in production.
  validates :password, not_pwned: { on_error: :raise_error }

  # Call custom proc on error. For example, capture errors in Sentry,
  # but do not mark the record as invalid.
  validates :password, not_pwned: {
    on_error: ->(record, error) { Raven.capture_exception(error) }
  }
end

You can configure network requests made from the validator using :request_options (see Net::HTTP.start for the list of available options).

  validates :password, not_pwned: {
    request_options: {
      read_timeout: 5,
      open_timeout: 1
    }
  }

These options override the globally defined default options (see above).

In addition to these options, you can also set the following:

HTTP headers can be specified with the :headers key (e.g. "User-Agent")

  validates :password, not_pwned: {
    request_options: {
      headers: { "User-Agent" => "Super fun user agent" }
    }
  }

An HTTP proxy can be set using the http_proxy or HTTP_PROXY environment variable. This is the same way that Net::HTTP handles HTTP proxies if no proxy options are given. See URI::Generic#find_proxy for full details on how Ruby detects a proxy from the environment.

  # Set in the environment
  ENV["http_proxy"] = "https://username:password@example.com:12345"

  validates :password, not_pwned: true

You can specify a custom HTTP proxy with the :proxy key:

  validates :password, not_pwned: {
    request_options: {
      proxy: "https://username:password@example.com:12345"
    }
  }

If you don't want to set a proxy and you don't want a proxy to be inferred from the environment, set the :ignore_env_proxy key:

  validates :password, not_pwned: {
    request_options: {
      ignore_env_proxy: true
    }
  }

You may have a use case for hashing the password in advance, and then making the call to the Pwned Passwords API later (for example if you want to enqueue a job without storing the plaintext password). To do this, you can hash the password with the Pwned.hash_password method and then initialize the Pwned::HashedPassword class with the hash, like this:

hashed_password = Pwned.hash_password(password)
# some time later
Pwned::HashedPassword.new(hashed_password, request_options).pwned?

The Pwned::HashedPassword constructor takes all the same options as the regular Pwned::Password constructor.

If you are using Devise I recommend you use the devise-pwned_password extension which is now powered by this gem.

If you are using Rodauth then you can use the rodauth-pwned feature which is powered by this gem.

The gem provides a command line utility for checking passwords. You can call it from your terminal application like this:

$ pwned password
Pwned!
The password has been found in public breaches 3645804 times.

If you don't want the password you are checking to be visible, call:

$ pwned --secret

You will be prompted for the password, but it won't be displayed.

To cut down on unnecessary network requests, the unpwn project uses a list of the top one million passwords to check passwords against. Only if a password is not included in the top million is it then checked against the Pwned Passwords API.

@daz shared a fantastic example of using this gem to show how many times the digits of Pi have been used as passwords and leaked.

require 'pwned'

PI = '3.14159265358979323846264338327950288419716939937510582097494459230781640628620899862803482534211706798214808651328230664709384460955058223172535940812848111'

for n in 1..40
  password = Pwned::Password.new PI[0..(n + 1)]
  str = [ n.to_s.rjust(2) ]
  str << (password.pwned? ? '😡' : '😃')
  str << password.pwned_count.to_s.rjust(4)
  str << password.password

  puts str.join ' '
end

The results may, or may not, surprise you.

 1 😡   16 3.1
 2 😡  238 3.14
 3 😡   34 3.141
 4 😡 1345 3.1415
 5 😡 2552 3.14159
 6 😡  791 3.141592
 7 😡 9582 3.1415926
 8 😡 1591 3.14159265
 9 😡  637 3.141592653
10 😡  873 3.1415926535
11 😡  137 3.14159265358
12 😡  103 3.141592653589
13 😡   65 3.1415926535897
14 😡  201 3.14159265358979
15 😡   41 3.141592653589793
16 😡   57 3.1415926535897932
17 😡   28 3.14159265358979323
18 😡   29 3.141592653589793238
19 😡    1 3.1415926535897932384
20 😡    7 3.14159265358979323846
21 😡    5 3.141592653589793238462
22 😡    2 3.1415926535897932384626
23 😡    2 3.14159265358979323846264
24 😃    0 3.141592653589793238462643
25 😡    3 3.1415926535897932384626433
26 😃    0 3.14159265358979323846264338
27 😃    0 3.141592653589793238462643383
28 😃    0 3.1415926535897932384626433832
29 😃    0 3.14159265358979323846264338327
30 😃    0 3.141592653589793238462643383279
31 😃    0 3.1415926535897932384626433832795
32 😃    0 3.14159265358979323846264338327950
33 😃    0 3.141592653589793238462643383279502
34 😃    0 3.1415926535897932384626433832795028
35 😃    0 3.14159265358979323846264338327950288
36 😃    0 3.141592653589793238462643383279502884
37 😃    0 3.1415926535897932384626433832795028841
38 😃    0 3.14159265358979323846264338327950288419
39 😃    0 3.141592653589793238462643383279502884197
40 😃    0 3.1415926535897932384626433832795028841971

After checking out the repository, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Bug reports and pull requests are welcome on GitHub at https://github.com/philnash/pwned. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.

The gem is available as open source under the terms of the MIT License.

Everyone interacting in the Pwned project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.