diff --git a/config/config.exs b/config/config.exs
index 89a34e0b..7805d7c5 100644
--- a/config/config.exs
+++ b/config/config.exs
@@ -29,48 +29,40 @@ use Mix.Config
 #
 #     import_config "#{Mix.env}.exs"
 
-config :crawly, Crawly.Worker, client: HTTPoison
-
 config :crawly,
   fetcher: {Crawly.Fetchers.HTTPoisonFetcher, []},
-  retry:
-    [
-      retry_codes: [400],
-      max_retries: 3,
-      ignored_middlewares: [Crawly.Middlewares.UniqueRequest]
+  retry: [
+    retry_codes: [400],
+    max_retries: 3,
+    ignored_middlewares: [Crawly.Middlewares.UniqueRequest]
   ],
 
-  # User agents which are going to be used with requests
-  user_agents: [
-    "Mozilla/5.0 (Macintosh; Intel Mac OS X x.y; rv:42.0) Gecko/20100101 Firefox/42.0",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41"
-  ],
-  # Item definition
-  item: [:title, :author, :time, :url],
-  # Identifier which is used to filter out duplicates
-  item_id: :title,
   # Stop spider after scraping certain amount of items
   closespider_itemcount: 500,
   # Stop spider if it does crawl fast enough
   closespider_timeout: 20,
   concurrent_requests_per_domain: 5,
+
+  # TODO: this looks outdated
   follow_redirect: true,
+
   # Request middlewares
   middlewares: [
     Crawly.Middlewares.DomainFilter,
     Crawly.Middlewares.UniqueRequest,
     Crawly.Middlewares.RobotsTxt,
-    Crawly.Middlewares.UserAgent
+    {Crawly.Middlewares.UserAgent,
+     user_agents: [
+       "Mozilla/5.0 (Macintosh; Intel Mac OS X x.y; rv:42.0) Gecko/20100101 Firefox/42.0",
+       "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
+       "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41"
+     ]}
   ],
   pipelines: [
-    Crawly.Pipelines.Validate,
-    Crawly.Pipelines.DuplicatesFilter,
+    {Crawly.Pipelines.Validate, fields: [:title, :author, :time, :url]},
+    {Crawly.Pipelines.DuplicatesFilter, item_id: :title},
     Crawly.Pipelines.JSONEncoder
   ]
 
-config :crawly, Crawly.Pipelines.WriteToFile,
-  folder: "/tmp",
-  extension: "jl"
 
- import_config "#{Mix.env}.exs"
+import_config "#{Mix.env()}.exs"
diff --git a/config/test.exs b/config/test.exs
index 6a006407..ad9c1d78 100644
--- a/config/test.exs
+++ b/config/test.exs
@@ -2,28 +2,30 @@ use Mix.Config
 
 config :crawly,
        manager_operations_timeout: 30_000,
-         # User agents which are going to be used with requests
-       user_agents: [
-         "Mozilla/5.0 (Macintosh; Intel Mac OS X x.y; rv:42.0) Gecko/20100101 Firefox/42.0",
-         "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
-         "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41"
-       ],
+
          # Stop spider after scraping certain amount of items
        closespider_itemcount: 100,
          # Stop spider if it does crawl fast enough
        closespider_timeout: 20,
        concurrent_requests_per_domain: 5,
        follow_redirect: true,
-         # Request middlewares
+
+       # Request middlewares
+       # User agents which are going to be used with requests
+       user_agents: [
+         "Mozilla/5.0 (Macintosh; Intel Mac OS X x.y; rv:42.0) Gecko/20100101 Firefox/42.0",
+         "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
+         "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41"
+       ],
        middlewares: [
          Crawly.Middlewares.DomainFilter,
          Crawly.Middlewares.UniqueRequest,
          Crawly.Middlewares.RobotsTxt,
-         Crawly.Middlewares.UserAgent
+         {Crawly.Middlewares.UserAgent, user_agents: ["My Custom Bot"]}
        ],
        pipelines: [
-         Crawly.Pipelines.Validate,
-         Crawly.Pipelines.DuplicatesFilter,
+         {Crawly.Pipelines.Validate, fields: [:title, :url, :time, :author]},
+         {Crawly.Pipelines.DuplicatesFilter, item_id: :title},
          Crawly.Pipelines.JSONEncoder
        ],
        retry: [
diff --git a/documentation/configuration.md b/documentation/configuration.md
index b98064b4..6b87f472 100644
--- a/documentation/configuration.md
+++ b/documentation/configuration.md
@@ -16,49 +16,6 @@ config :crawly,
 
 ## Options
 
-### base_store_path :: binary() [DEPRECATED in 0.6.0]
-
-default: "/tmp"
-
-Defines the path where items are stored in the filesystem. This setting
-is used by the Crawly.DataStorageWorker process.
-
-> **Deprecated**: This has been deprecated in favour of having pipelines to handle data storage, as of `0.6.0`
-
-### `user_agents` :: list()
-
-default: ["Crawly Bot 1.0"]
-
-Defines a user agent string for Crawly requests. This setting is used
-by the `Crawly.Middlewares.UserAgent` middleware. When the list has more than one
-item, all requests will be executed, each with a user agent string chosen
-randomly from the supplied list.
-
-> **Deprecated**: This has been deprecated in favour of tuple-based pipeline configuration instead of global configurations, as of `0.7.0`. Refer to `Crawly.Middlewares.UserAgent` module documentation for correct usage.
-
-### `item` :: [atom()]
-
-default: []
-
-Defines a list of required fields for the item. When none of the default
-fields are added to the following item (or if the values of
-required fields are "" or nil), the item will be dropped. This setting
-is used by the `Crawly.Pipelines.Validate` pipeline
-
-> **Deprecated**: This has been deprecated in favour of tuple-based pipeline configuration instead of global configurations, as of `0.7.0`. Refer to `Crawly.Pipelines.Validate` module documentation for correct usage.
-
-### `item_id` :: atom()
-
-default: nil
-
-Defines a field which will be used in order to identify if an item is
-a duplicate or not. In most of the ecommerce websites the desired id
-field is the SKU. This setting is used in
-the `Crawly.Pipelines.DuplicatesFilter` pipeline. If unset, the related
-middleware is effectively disabled.
-
-> **Deprecated**: This has been deprecated in favour of tuple-based pipeline configuration instead of global configurations, as of `0.7.0`. Refer to `Crawly.Pipelines.DuplicatesFilter` module documentation for correct usage.
-
 ### `pipelines` :: [module()]
 
 default: []
@@ -99,17 +56,12 @@ default: :disabled
 
 An integer which specifies a number of items. If the spider scrapes more than that amount and those items are passed by the item pipeline, the spider will be closed. If set to :disabled the spider will not be stopped.
 
-### closespider_timeout :: pos_integer()
+### closespider_timeout :: pos_integer() | :disabled
 
 default: nil
 
 Defines a minimal amount of items which needs to be scraped by the spider within the given timeframe (30s). If the limit is not reached by the spider - it will be stopped.
 
-### follow_redirect :: boolean()
-
-default: false
-
-Defines is Crawly spider is supposed to follow HTTP redirects or not.
 
 ### concurrent_requests_per_domain :: pos_integer()
 
diff --git a/lib/crawly/data_storage/data_storage_worker.ex b/lib/crawly/data_storage/data_storage_worker.ex
index 3f9bfcf9..35d53f29 100644
--- a/lib/crawly/data_storage/data_storage_worker.ex
+++ b/lib/crawly/data_storage/data_storage_worker.ex
@@ -33,7 +33,7 @@ defmodule Crawly.DataStorage.Worker do
   end
 
   def handle_cast({:store, item}, state) do
-    pipelines = Application.get_env(:crawly, :pipelines, [])
+    pipelines = Crawly.Utils.get_settings(:pipelines, state.spider_name, [])
 
     state =
       case Crawly.Utils.pipe(pipelines, item, state) do
diff --git a/lib/crawly/fetchers/fetcher.ex b/lib/crawly/fetchers/fetcher.ex
index 20a96b13..49e4b4ec 100644
--- a/lib/crawly/fetchers/fetcher.ex
+++ b/lib/crawly/fetchers/fetcher.ex
@@ -6,6 +6,8 @@ defmodule Crawly.Fetchers.Fetcher do
   Crawly.Request, HTTP client options and return Crawly.Response.
   """
 
+  @type t :: {module(), list()}
+
   @callback fetch(request, options) :: {:ok, response} | {:error, reason}
             when request: Crawly.Request.t(),
                  response: Crawly.Response.t(),
diff --git a/lib/crawly/manager.ex b/lib/crawly/manager.ex
index 37b25b0f..dcb5a0db 100644
--- a/lib/crawly/manager.ex
+++ b/lib/crawly/manager.ex
@@ -31,6 +31,8 @@ defmodule Crawly.Manager do
 
   use GenServer
 
+  alias Crawly.Utils
+
   def start_link(spider_name) do
     Logger.debug("Starting the manager for #{spider_name}")
     GenServer.start_link(__MODULE__, spider_name)
@@ -57,7 +59,7 @@ defmodule Crawly.Manager do
 
     # Start workers
     num_workers =
-      Application.get_env(:crawly, :concurrent_requests_per_domain, 4)
+      Utils.get_settings(:concurrent_requests_per_domain, spider_name, 4)
 
     worker_pids =
       Enum.map(1..num_workers, fn _x ->
@@ -72,8 +74,15 @@ defmodule Crawly.Manager do
     )
 
     # Schedule basic service operations for given spider manager
-    tref = Process.send_after(self(), :operations, get_timeout())
-    {:ok, %{name: spider_name, tref: tref, prev_scraped_cnt: 0}}
+    tref =
+      Process.send_after(
+        self(),
+        :operations,
+        Utils.get_settings(:manager_operations_timeout, spider_name, @timeout)
+      )
+
+    {:ok,
+     %{name: spider_name, tref: tref, prev_scraped_cnt: 0, workers: worker_pids}}
   end
 
   def handle_info(:operations, state) do
@@ -85,7 +94,7 @@ defmodule Crawly.Manager do
     delta = items_count - state.prev_scraped_cnt
     Logger.info("Current crawl speed is: #{delta} items/min")
 
-    case Application.get_env(:crawly, :closespider_itemcount, :disabled) do
+    case Utils.get_settings(:closespider_itemcount, state.name, :disabled) do
       :disabled ->
         :ignored
 
@@ -100,8 +109,8 @@ defmodule Crawly.Manager do
         :ignoring
     end
 
-    # Close spider in case if it's not scraping itms fast enough
-    case Application.get_env(:crawly, :closespider_timeout) do
+    # Close spider in case if it's not scraping items fast enough
+    case Utils.get_settings(:closespider_timeout, state.name, :disabled) do
       :undefined ->
         :ignoring
 
@@ -116,12 +125,13 @@ defmodule Crawly.Manager do
         :ignoring
     end
 
-    tref = Process.send_after(self(), :operations, get_timeout())
+    tref =
+      Process.send_after(
+        self(),
+        :operations,
+        Utils.get_settings(:manager_operations_timeout, state.name, @timeout)
+      )
 
     {:noreply, %{state | tref: tref, prev_scraped_cnt: items_count}}
   end
-
-  defp get_timeout() do
-    Application.get_env(:crawly, :manager_operations_timeout, @timeout)
-  end
 end
diff --git a/lib/crawly/middlewares/user_agent.ex b/lib/crawly/middlewares/user_agent.ex
index fe897a18..fa51779e 100644
--- a/lib/crawly/middlewares/user_agent.ex
+++ b/lib/crawly/middlewares/user_agent.ex
@@ -12,7 +12,7 @@ defmodule Crawly.Middlewares.UserAgent do
   ### Example Declaration
   ```
   middlewares: [
-    {UserAgent, user_agents: ["My Custom Bot] }
+    {UserAgent, user_agents: ["My Custom Bot"] }
   ]
   ```
   """
@@ -24,8 +24,7 @@ defmodule Crawly.Middlewares.UserAgent do
     new_headers = List.keydelete(request.headers, "User-Agent", 0)
 
     user_agents =
-      Map.get(opts, :user_agents) ||
-        Application.get_env(:crawly, :user_agents, ["Crawly Bot 1.0"])
+      Map.get(opts, :user_agents, ["Crawly Bot 1.0"])
 
     useragent = Enum.random(user_agents)
 
diff --git a/lib/crawly/pipelines/csv_encoder.ex b/lib/crawly/pipelines/csv_encoder.ex
index f4af184c..a2c00b62 100644
--- a/lib/crawly/pipelines/csv_encoder.ex
+++ b/lib/crawly/pipelines/csv_encoder.ex
@@ -18,7 +18,7 @@ defmodule Crawly.Pipelines.CSVEncoder do
           {false, state :: map} | {csv_line :: String.t(), state :: map}
   def run(item, state, opts \\ []) do
     opts = Enum.into(opts, %{fields: nil})
-    fields = Map.get(opts, :fields) || Application.get_env(:crawly, :item)
+    fields = Map.get(opts, :fields, [])
 
     case fields do
       :undefined ->
diff --git a/lib/crawly/pipelines/duplicates_filter.ex b/lib/crawly/pipelines/duplicates_filter.ex
index 3c0a2003..6aecdb9e 100644
--- a/lib/crawly/pipelines/duplicates_filter.ex
+++ b/lib/crawly/pipelines/duplicates_filter.ex
@@ -32,7 +32,7 @@ defmodule Crawly.Pipelines.DuplicatesFilter do
   def run(item, state, opts \\ []) do
     opts = Enum.into(opts, %{item_id: nil})
 
-    item_id = Map.get(opts, :item_id) || Application.get_env(:crawly, :item_id)
+    item_id = Map.get(opts, :item_id)
 
     item_id = Map.get(item, item_id)
 
diff --git a/lib/crawly/pipelines/validate.ex b/lib/crawly/pipelines/validate.ex
index 711dd8ad..591d3045 100644
--- a/lib/crawly/pipelines/validate.ex
+++ b/lib/crawly/pipelines/validate.ex
@@ -27,7 +27,7 @@ defmodule Crawly.Pipelines.Validate do
   @impl Crawly.Pipeline
   def run(item, state, opts \\ []) do
     opts = Enum.into(opts, %{fields: nil})
-    fields = Map.get(opts, :fields) || Application.get_env(:crawly, :item, [])
+    fields = Map.get(opts, :fields, [])
 
     validation_result =
       fields
diff --git a/lib/crawly/pipelines/write_to_file.ex b/lib/crawly/pipelines/write_to_file.ex
index b451bf65..95761175 100644
--- a/lib/crawly/pipelines/write_to_file.ex
+++ b/lib/crawly/pipelines/write_to_file.ex
@@ -59,20 +59,11 @@ defmodule Crawly.Pipelines.WriteToFile do
   def run(item, state, opts) do
     opts = Enum.into(opts, %{folder: nil, extension: nil})
 
-    global_config =
-      Application.get_env(
-        :crawly,
-        Crawly.Pipelines.WriteToFile,
-        Keyword.new()
-      )
-
     folder =
-      Map.get(opts, :folder) ||
-        Keyword.get(global_config, :folder, System.tmp_dir!())
+      Map.get(opts, :folder, "./")
 
     extension =
-      Map.get(opts, :extension) ||
-        Keyword.get(global_config, :extension, "jl")
+      Map.get(opts, :extension, "jl")
 
     fd = open_fd(state.spider_name, folder, extension)
     :ok = write(fd, item)
diff --git a/lib/crawly/settings.ex b/lib/crawly/settings.ex
new file mode 100644
index 00000000..30ba7bfa
--- /dev/null
+++ b/lib/crawly/settings.ex
@@ -0,0 +1,48 @@
+defmodule Crawly.Settings do
+  @moduledoc """
+  Define Crawly setting types
+  """
+
+  @type numeric_setting() :: pos_integer() | :disabled
+  @type retry() :: [
+          retry_codes: [pos_integer()],
+          max_retries: pos_integer(),
+          ignored_middlewares: [module()]
+        ]
+
+  @type middleware() ::
+          Crawly.Middlewares.DomainFilter
+          | Crawly.Middlewares.UniqueRequest
+          | Crawly.Middlewares.RobotsTxt
+          | Crawly.Middlewares.AutoCookiesManager
+          | {Crawly.Middlewares.UserAgent, user_agents: [binary()]}
+
+  @type pipeline() ::
+          Crawly.Pipelines.JSONEncoder
+          | {Crawly.Pipelines.DuplicatesFilter, item_id: atom()}
+          | {Crawly.Pipelines.Validate, fields: [atom()]}
+          | {Crawly.Pipelines.CSVEncoder, fields: [atom()]}
+          | {Crawly.Pipelines.WriteToFile,
+             folder: binary(), extension: binary()}
+
+  @type t() :: [
+          # Allows to stop spider after a given number of scraped items
+          # :disabled by default.
+          closespider_itemcount: numeric_setting(),
+
+          # Allows to stop spider if it extracts less than a given amount of
+          # items per minute.
+          closespider_timeout: pos_integer(),
+
+          # Allows to control how many workers are started for a given domain
+          concurrent_requests_per_domain: pos_integer(),
+
+          # Allows to define a fetcher to perform HTTP requests
+          fetcher: Crawly.Fetchers.Fetcher.t(),
+
+          # Defines retries
+          retry: retry(),
+          middlewares: [middleware()],
+          pipelines: [pipeline()]
+        ]
+end
diff --git a/lib/crawly/spider.ex b/lib/crawly/spider.ex
index 0be41baf..e38c4a4d 100644
--- a/lib/crawly/spider.ex
+++ b/lib/crawly/spider.ex
@@ -9,6 +9,10 @@ defmodule Crawly.Spider do
   3. `parse_item/1` function which is responsible for parsing the downloaded
      request and converting it into items which can be stored and new requests
      which can be scheduled
+  4. `custom_settings/0` an optional callback which can be used in order to
+      provide custom spider specific settings. Should define a list with custom
+      settings and their values. These values will take precedence over the
+      global settings defined in the config.
   """
 
   @callback init() :: [start_urls: list()]
@@ -16,6 +20,9 @@ defmodule Crawly.Spider do
   @callback base_url() :: binary()
 
   @callback parse_item(response :: HTTPoison.Response.t()) ::
-  Crawly.ParsedItem.t()
+              Crawly.ParsedItem.t()
 
+  @callback override_settings() :: Crawly.Settings.t()
+
+  @optional_callbacks override_settings: 0
 end
diff --git a/lib/crawly/utils.ex b/lib/crawly/utils.ex
index 8a94de3e..86b8f86e 100644
--- a/lib/crawly/utils.ex
+++ b/lib/crawly/utils.ex
@@ -93,9 +93,9 @@ defmodule Crawly.Utils do
       catch
         error, reason ->
           Logger.error(
-            "Pipeline crash: #{module}, error: #{inspect(error)}, reason: #{
-              inspect(reason)
-            }, args: #{inspect(args)}"
+            "Pipeline crash: #{module}, error: #{inspect(error)}, reason: #{inspect(reason)}, args: #{
+              inspect(args)
+            }"
           )
 
           {item, state}
@@ -113,4 +113,50 @@ defmodule Crawly.Utils do
   def send_after(pid, message, timeout) do
     Process.send_after(pid, message, timeout)
   end
+
+  @doc """
+  A helper which allows to extract a given setting.
+
+  Returned value is a result of intersection of the global settings and settings
+  defined as settings_override inside the spider. Settings defined on spider are
+  taking precedence over the global settings defined in the config.
+  """
+  @spec get_settings(setting_name, spider_name, default) :: result
+    when setting_name: atom(),
+         spider_name: atom(),
+         default: term(),
+         result: term()
+
+  def get_settings(setting_name, spider_name \\ nil, default \\ nil) do
+    global_setting = Application.get_env(:crawly, setting_name, default)
+    case get_spider_setting(setting_name, spider_name) do
+      nil ->
+        # No custom settings for a spider found
+        global_setting
+
+      custom_setting ->
+        custom_setting
+    end
+  end
+
+  ##############################################################################
+  # Private functions
+  ##############################################################################
+  @spec get_spider_setting(spider_name, setting_name) :: result
+        when spider_name: atom(),
+             setting_name: atom(),
+             result: nil | term()
+
+  defp get_spider_setting(_setting_name, nil), do: nil
+
+  defp get_spider_setting(setting_name, spider_name) do
+    case function_exported?(spider_name, :override_settings, 0) do
+      true ->
+
+        Keyword.get(spider_name.override_settings(), setting_name, nil)
+
+      false ->
+        nil
+    end
+  end
 end
diff --git a/lib/crawly/worker.ex b/lib/crawly/worker.ex
index af6cb5ac..addc542c 100644
--- a/lib/crawly/worker.ex
+++ b/lib/crawly/worker.ex
@@ -42,15 +42,12 @@ defmodule Crawly.Worker do
 
           case :epipe.run(functions, {request, spider_name}) do
             {:error, _step, reason, _step_state} ->
-
-              Logger.debug(
-                fn ->
-                  "Crawly worker could not process the request to #{
-                    inspect(request.url)
-                  }
+              Logger.debug(fn ->
+                "Crawly worker could not process the request to #{
+                  inspect(request.url)
+                }
                   reason: #{inspect(reason)}"
-                end
-              )
+              end)
 
               @default_backoff
 
@@ -73,14 +70,16 @@ defmodule Crawly.Worker do
     # check if spider-level fetcher is set. Overrides the globally configured fetcher.
     # if not set, log warning for explicit config preferred,
     # get the globally-configured fetcher. Defaults to HTTPoisonFetcher
-    {fetcher, options} = Application.get_env(
-      :crawly,
-      :fetcher,
-      {Crawly.Fetchers.HTTPoisonFetcher, []}
-    )
-
-    retry_options = Application.get_env(:crawly, :retry, [])
+    {fetcher, options} =
+      Crawly.Utils.get_settings(
+        :fetcher,
+        spider_name,
+        {Crawly.Fetchers.HTTPoisonFetcher, []}
+      )
+
+    retry_options = Crawly.Utils.get_settings(:retry, spider_name, [])
     retry_codes = Keyword.get(retry_options, :retry_codes, [])
+
     case fetcher.fetch(request, options) do
       {:error, _reason} = err ->
         :ok = maybe_retry_request(spider_name, request)
@@ -93,11 +92,11 @@ defmodule Crawly.Worker do
           true ->
             :ok = maybe_retry_request(spider_name, request)
             {:error, :retry}
+
           false ->
             {:ok, {response, spider_name}}
         end
     end
-
   end
 
   @spec parse_item({response, spider_name}) :: result
@@ -141,7 +140,8 @@ defmodule Crawly.Worker do
       fn request ->
         request = Map.put(request, :prev_response, response)
         Crawly.RequestsStorage.store(spider_name, request)
-      end)
+      end
+    )
 
     # Process all items one by one
     Enum.each(
@@ -157,7 +157,7 @@ defmodule Crawly.Worker do
   ## Retry a request if max retries allows to do so
   defp maybe_retry_request(spider, request) do
     retries = request.retries
-    retry_settings = Application.get_env(:crawly, :retry, Keyword.new())
+    retry_settings = Crawly.Utils.get_settings(:retry, spider, Keyword.new())
 
     ignored_middlewares = Keyword.get(retry_settings, :ignored_middlewares, [])
     max_retries = Keyword.get(retry_settings, :max_retries, 0)
@@ -169,16 +169,16 @@ defmodule Crawly.Worker do
         middlewares = request.middlewares -- ignored_middlewares
 
         request = %Crawly.Request{
-          request |
-          middlewares: middlewares,
-          retries: retries + 1
+          request
+          | middlewares: middlewares,
+            retries: retries + 1
         }
 
         :ok = Crawly.RequestsStorage.store(spider, request)
+
       false ->
         Logger.error("Dropping request to #{request.url}, (max retries)")
         :ok
     end
-
   end
 end
diff --git a/test/data_storage_test.exs b/test/data_storage_test.exs
index a65aff05..fe94bae8 100644
--- a/test/data_storage_test.exs
+++ b/test/data_storage_test.exs
@@ -73,30 +73,4 @@ defmodule DataStorageTest do
     result = Crawly.DataStorage.stats(:unkown)
     assert result == {:error, :data_storage_worker_not_running}
   end
-
-  test "Duplicates pipline is inactive when item_id is not set", context do
-    :meck.expect(Application, :get_env, fn :crawly, :item_id -> :undefined end)
-
-    Crawly.DataStorage.store(context.crawler, %{
-      title: "test title",
-      author: "me",
-      time: "Now",
-      url: "http://example.com"
-    })
-
-    Crawly.DataStorage.store(context.crawler, %{
-      title: "test title",
-      author: "me",
-      time: "Now",
-      url: "http://example.com"
-    })
-
-    Process.sleep(1000)
-    {:stored_items, 2} = Crawly.DataStorage.stats(context.crawler)
-    :meck.unload(Application)
-  end
-
-
-#  test "No data is stored if Dev"
-
 end
diff --git a/test/manager_test.exs b/test/manager_test.exs
index 204ecdde..9fb61fce 100644
--- a/test/manager_test.exs
+++ b/test/manager_test.exs
@@ -54,8 +54,6 @@ defmodule ManagerTest do
     Application.put_env(:crawly, :concurrent_requests_per_domain, 1)
     :ok = Crawly.Engine.start_spider(Manager.TestSpider)
     Process.sleep(2_000)
-    stats = Crawly.DataStorage.stats(Manager.TestSpider)
-    IO.puts("Stats: #{inspect(stats)}")
     assert %{} == Crawly.Engine.running_spiders()
   end
 
diff --git a/test/middlewares/user_agent_test.exs b/test/middlewares/user_agent_test.exs
index b07cd116..5d5133e0 100644
--- a/test/middlewares/user_agent_test.exs
+++ b/test/middlewares/user_agent_test.exs
@@ -1,15 +1,8 @@
 defmodule Middlewares.UserAgentTest do
   use ExUnit.Case, async: false
 
-  setup do
-    on_exit(fn ->
-      Application.put_env(:crawly, :user_agents, nil)
-    end)
-  end
-
   test "Adds a user agent to request header with global config" do
-    Application.put_env(:crawly, :user_agents, ["My Custom Bot"])
-    middlewares = [Crawly.Middlewares.UserAgent]
+    middlewares = [{Crawly.Middlewares.UserAgent, user_agents: ["My Custom Bot"]}]
     req = %Crawly.Request{}
     state = %{}
 
diff --git a/test/pipelines/csv_encoder_test.exs b/test/pipelines/csv_encoder_test.exs
index ac660a65..a73dfe6f 100644
--- a/test/pipelines/csv_encoder_test.exs
+++ b/test/pipelines/csv_encoder_test.exs
@@ -2,16 +2,8 @@ defmodule Pipelines.CSVEncoderTest do
   use ExUnit.Case, async: false
 
   @valid %{first: "some", second: "data"}
-  setup do
-    on_exit(fn ->
-      Application.put_env(:crawly, :item, [:title, :author, :time, :url])
-    end)
-  end
-
   test "Converts a single-level map to a csv string with global config" do
-    Application.put_env(:crawly, :item, [:first, :second])
-
-    pipelines = [Crawly.Pipelines.CSVEncoder]
+    pipelines = [{Crawly.Pipelines.CSVEncoder, fields: [:first, :second]}]
     item = @valid
     state = %{}
 
diff --git a/test/pipelines/duplicates_filter_test.exs b/test/pipelines/duplicates_filter_test.exs
index 383120de..883cfd48 100644
--- a/test/pipelines/duplicates_filter_test.exs
+++ b/test/pipelines/duplicates_filter_test.exs
@@ -2,15 +2,9 @@ defmodule Pipelines.DuplicatesFilterTest do
   use ExUnit.Case, async: false
 
   @valid %{data: [%{some: "nested_data"}], id: "my_id"}
-  setup do
-    on_exit(fn ->
-      Application.put_env(:crawly, :item_id, :title)
-    end)
-  end
 
   test "Drops duplicate items with the same item_id value through global config" do
-    Application.put_env(:crawly, :item_id, :id)
-    pipelines = [Crawly.Pipelines.DuplicatesFilter]
+    pipelines = [{Crawly.Pipelines.DuplicatesFilter, item_id: :id}]
     item = @valid
     state = %{}
 
diff --git a/test/pipelines/validate_test.exs b/test/pipelines/validate_test.exs
index 49671eaa..e948a09a 100644
--- a/test/pipelines/validate_test.exs
+++ b/test/pipelines/validate_test.exs
@@ -13,15 +13,9 @@ defmodule Pipelines.ValidateTest do
     author: nil
   }
 
-  setup do
-    on_exit(fn ->
-      Application.put_env(:crawly, :item, [:title, :author, :time, :url])
-    end)
-  end
-
   test "Returns item unchanged when has required fields" do
     Application.put_env(:crawly, :item, [:title, :author])
-    pipelines = [Crawly.Pipelines.Validate]
+    pipelines = [{Crawly.Pipelines.Validate, fields: [:title, :author]}]
     item = @valid
     state = %{}
 
@@ -30,8 +24,7 @@ defmodule Pipelines.ValidateTest do
   end
 
   test "Drops items when missing required fields with global config" do
-    Application.put_env(:crawly, :item, [:title, :author])
-    pipelines = [Crawly.Pipelines.Validate]
+    pipelines = [{Crawly.Pipelines.Validate, fields: [:title, :author]}]
     item = @invalid_missing
     state = %{}
 
@@ -47,8 +40,7 @@ defmodule Pipelines.ValidateTest do
   end
 
   test "Drops items when required fields are equal to nil" do
-    Application.put_env(:crawly, :item, [:title, :author])
-    pipelines = [Crawly.Pipelines.Validate]
+    pipelines = [{Crawly.Pipelines.Validate, fields: [:title, :author]}]
     item = @invalid_nil
     state = %{}
 
diff --git a/test/pipelines/write_to_file_test.exs b/test/pipelines/write_to_file_test.exs
index 3f1a2883..fc249038 100644
--- a/test/pipelines/write_to_file_test.exs
+++ b/test/pipelines/write_to_file_test.exs
@@ -5,7 +5,6 @@ defmodule Pipelines.WriteToFileTest do
 
   setup do
     on_exit(fn ->
-      Application.put_env(:crawly, :"Crawly.Pipelines.WriteToFile", nil)
       :meck.unload(IO)
       :meck.unload(File)
     end)
@@ -31,15 +30,8 @@ defmodule Pipelines.WriteToFileTest do
       end
     )
 
-    Application.put_env(
-      :crawly,
-      Crawly.Pipelines.WriteToFile,
-      folder: "/tmp",
-      extension: "csv"
-    )
-
     pipelines = [
-      Crawly.Pipelines.WriteToFile
+      {Crawly.Pipelines.WriteToFile, folder: "/tmp", extension: "csv"}
     ]
 
     item = @binary
diff --git a/test/settings_test.exs b/test/settings_test.exs
new file mode 100644
index 00000000..5a15f792
--- /dev/null
+++ b/test/settings_test.exs
@@ -0,0 +1,46 @@
+defmodule SettingsTest do
+  use ExUnit.Case
+
+  setup do
+    Application.ensure_all_started(Crawly)
+    Application.put_env(:crawly, :concurrent_requests_per_domain, 1)
+    Application.put_env(:crawly, :closespider_itemcount, 10)
+
+    on_exit(fn ->
+      Application.put_env(:crawly, :closespider_timeout, 20)
+      Application.put_env(:crawly, :closespider_itemcount, 100)
+    end)
+  end
+
+  test "settings from the spider are overriding globals" do
+    assert 5 ==
+             Crawly.Utils.get_settings(
+               :concurrent_requests_per_domain,
+               TestSpiderSettingsOverride, 1)
+  end
+
+  test "incomplete spider overrides do not break global settings" do
+    assert 10 ==
+             Crawly.Utils.get_settings(
+               :closespider_itemcount,
+               TestSpiderSettingsOverride, 1)
+  end
+end
+
+defmodule Elixir.TestSpiderSettingsOverride do
+  def base_url() do
+    "https://www.example.com"
+  end
+
+  def init() do
+    [
+      start_urls: ["https://www.example.com/blog.html"]
+    ]
+  end
+
+  def parse_item(_response) do
+    %{:items => [], :requests => []}
+  end
+
+  def override_settings(), do: [concurrent_requests_per_domain: 5]
+end
diff --git a/test/utils_test.exs b/test/utils_test.exs
index b47c09de..4cea2f93 100644
--- a/test/utils_test.exs
+++ b/test/utils_test.exs
@@ -100,7 +100,7 @@ defmodule UtilsTest do
         Crawly.Middlewares.DomainFilter,
         Crawly.Middlewares.UniqueRequest,
         Crawly.Middlewares.RobotsTxt,
-        Crawly.Middlewares.UserAgent],
+        {Crawly.Middlewares.UserAgent, user_agents: ["My Custom Bot"]}],
       retries: 0
     }
   end