Tokens provide a way to generate and verify bearer tokens for use in Channels or API authentication.

The data stored in the token is signed to prevent tampering but not encrypted. This means it is safe to store identification information (such as user IDs) but should not be used to store confidential information (such as credit card numbers).

Example

When generating a unique token for use in an API or Channel it is advised to use a unique identifier for the user, typically the id from a database. For example:

iex> user_id = 1
iex> token = Phoenix.Token.sign(MyApp.Endpoint, "user salt", user_id)
iex> Phoenix.Token.verify(MyApp.Endpoint, "user salt", token, max_age: 86400)
{:ok, 1}

In that example we have a user’s id, we generate a token and verify it using the secret key base configured in the given endpoint. We guarantee the token will only be valid for one day by setting a max age (recommended).

The first argument to both sign/4 and verify/4 can be one of:

• the module name of a Phoenix endpoint (shown above) - where the secret key base is extracted from the endpoint
• Plug.Conn - where the secret key base is extracted from the endpoint stored in the connection
• Phoenix.Socket - where the secret key base is extracted from the endpoint stored in the socket
• a string, representing the secret key base itself. A key base with at least 20 randomly generated characters should be used to provide adequate entropy

The second argument is a cryptographic salt which must be the same in both calls to sign/4 and verify/4. For instance, it may be called “user auth” when generating a token that will be used to authenticate users on channels or on your APIs.

The third argument can be any term (string, int, list, etc.) that you wish to codify into the token. Upon valid verification, this same term will be extracted from the token.

Usage

Once a token is signed, we can send it to the client in multiple ways.

One is via the meta tag:

<%= tag :meta, name: "channel_token",
               content: Phoenix.Token.sign(@conn, "user salt", @current_user.id) %>

Or an endpoint that returns it:

def create(conn, params) do
  user = User.create(params)
  render(conn, "user.json",
         %{token: Phoenix.Token.sign(conn, "user salt", user.id), user: user})
end

Once the token is sent, the client may now send it back to the server as an authentication mechanism. For example, we can use it to authenticate a user on a Phoenix channel:

defmodule MyApp.UserSocket do
  use Phoenix.Socket

  def connect(%{"token" => token}, socket) do
    case Phoenix.Token.verify(socket, "user salt", token, max_age: 86400) do
      {:ok, user_id} ->
        socket = assign(socket, :user, Repo.get!(User, user_id))
        {:ok, socket}
      {:error, _} ->
        :error
    end
  end
end

In this example, the phoenix.js client will send the token in the connect command which is then validated by the server.

Phoenix.Token can also be used for validating APIs, handling password resets, e-mail confirmation and more.