diff --git a/dff/pipeline/pipeline/actor.py b/dff/pipeline/pipeline/actor.py
index 1bd5878b0..56c902672 100644
--- a/dff/pipeline/pipeline/actor.py
+++ b/dff/pipeline/pipeline/actor.py
@@ -390,9 +390,7 @@ def validate_script(self, pipeline: Pipeline, verbose: bool = True):
         return error_msgs
 
 
-def default_condition_handler(
-    condition: Callable, ctx: Context, pipeline: Pipeline, *args, **kwargs
-) -> Callable[[Context, Pipeline, Any, Any], bool]:
+def default_condition_handler(condition: Callable, ctx: Context, pipeline: Pipeline, *args, **kwargs) -> bool:
     """
     The simplest and quickest condition handler for trivial condition handling returns the callable condition:
 
diff --git a/docs/source/user_guides.rst b/docs/source/user_guides.rst
index 0cb1a1531..a7f82f073 100644
--- a/docs/source/user_guides.rst
+++ b/docs/source/user_guides.rst
@@ -23,6 +23,14 @@ for exploring the telemetry data collected from your conversational services.
 We show how to plug in the telemetry collection and configure the pre-built
 Superset dashboard shipped with DFF.
 
+:doc:`User defined functions guide <./user_guides/user_functions>`
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The ``user defined functions guide`` tutorial describes useage of all user-defined functions
+that can be used for framework extension.
+For each of these functions use cases are defined, execution orser is specified
+and also signature and return value types are highlighted.
+
 :doc:`Optimization guide <./user_guides/optimization_guide>`
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
@@ -32,8 +40,6 @@ and to locate and remove performance bottlenecks.
 
 .. toctree::
    :hidden:
+   :glob:
 
-   user_guides/basic_conceptions
-   user_guides/context_guide
-   user_guides/superset_guide
-   user_guides/optimization_guide
+   user_guides/*
diff --git a/docs/source/user_guides/user_functions.rst b/docs/source/user_guides/user_functions.rst
new file mode 100644
index 000000000..b7e680322
--- /dev/null
+++ b/docs/source/user_guides/user_functions.rst
@@ -0,0 +1,468 @@
+User functions guide
+--------------------
+
+Overview
+++++++++
+
+Dialog flow framework allows user to define custom functions for implementing custom behavior
+in several aspects.
+This tutorial summarizes the custom functions use cases, specifies their arguments and return
+types, warns about several common exception handling cases.
+
+``Actor`` handlers
+++++++++++++++++++
+
+Description
+~~~~~~~~~~~
+
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ constructor accepts ``handlers``
+parameter, that is either ``None`` or dictionary attributing lists of functions to different
+`ActorStage <../api/dff.script.core.types#ActorStage>`__ values.
+
+These functions are run at specific point in `Actor <../api/dff.pipeline.pipeline.actor#Actor>`__
+lifecycle.
+
+Use-cases
+~~~~~~~~~
+
+TODO
+
+Signature
+~~~~~~~~~
+
+.. code-block:: python
+
+    def handler(ctx: Context, pipeline: Pipeline) -> Any:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value can be anything (it is not used).
+
+Exceptions
+~~~~~~~~~~
+
+If an exception occurs during this function execution, it will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the actor service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+These exceptions **are not raised** during script validation.
+
+Script functions
+++++++++++++++++
+
+Pre-transition and pre-response processors
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Description
+===========
+
+Each script `Node <../api/dff.script.core.script#Node>`__ has a properties called
+``pre_transitions_processing`` and ``pre_response_processing``.
+That is a dictionary, associating functions to their names (that can be any hashable object).
+
+Pre-transition proccessors are run before transition from the previous node to the current node.
+Pre-response proccessors are run before acquiring the response, after the current node is processed.
+
+Use-cases
+=========
+
+TODO
+
+See tutorials:
+
+#. `About pre-transition processors <../tutorials/tutorials.script.core.9_pre_transitions_processing>`__
+#. `About pre-response processors <../tutorials/tutorials.script.core.7_pre_response_processing>`__
+
+Signature
+=========
+
+.. code-block:: python
+
+    def handler(ctx: Context, pipeline: Pipeline) -> Context:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is the modified Context value.
+
+Exceptions
+==========
+
+If an exception occurs during this function execution, it will be handled internally,
+only an exception message will be printed to ``stdout``.
+These exceptions **are not raised** during script validation.
+
+Script conditions and condition handlers
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Description of condition
+========================
+
+Each transition between different nodes in `Script <../api/dff.script.core.script#Script>`__
+has a condition function.
+They are provided in script dictionary (TODO: parser?) in form of Python functions.
+
+These functions are executed on pipeline startup if ``validation_stage`` parameter of
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ is ``True`` and before transition
+from the corresponding node in script in runtime.
+
+Use-cases
+=========
+
+TODO
+
+See `tutorial about conditions <../tutorials/tutorials.script.core.2_conditions>`__.
+
+Signature of condition
+======================
+
+.. code-block:: python
+
+    def condition(ctx: Context, pipeline: Pipeline) -> bool:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is ``True`` if transition should be made and ``False`` otherwise.
+
+Standard conditions
+===================
+
+There is a set of `standard script condition functions <../api/dff.script.conditions.std_conditions>`__ defined.
+
+Exceptions in conditions
+========================
+
+If an exception occurs during this function execution, it will be reported during script validation stage
+(if any) and also will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the actor service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+
+Description of condition handler
+================================
+
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ constructor also accepts
+condition handler - that is a special function that executes conditions.
+
+This function is invoked every time condition should be checked, it launches and checks condition.
+
+Use-cases
+=========
+
+TODO
+
+Signature of condition handler
+==============================
+
+.. code-block:: python
+
+    def condition_handler(condition: Callable[[Context, Pipeline], bool], ctx: Context, pipeline: Pipeline) -> bool:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is ``True`` if transition should be made and ``False`` otherwise.
+
+Standard condition handler
+==========================
+
+The simplest `default condition handler <../api/dff.pipeline.pipeline.actor#default_condition_handler>`__
+just invokes the condition function and returns the result.
+
+Exceptions in condition handler
+================================
+
+If an exception occurs during this function execution, it will be reported during script validation stage
+(if any), otherwise it will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the actor service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+
+Labels
+~~~~~~
+
+Description
+===========
+
+Some of the transitions between nodes in `Script <../api/dff.script.core.script#Script>`__
+do not have "absolute" node targets specified.
+For instance, that might be useful in case it is required to stay in the same node or transition
+to the previous node.
+For such cases special function node labels can be used.
+
+These functions are executed on pipeline startup if ``validation_stage`` parameter of
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ is ``True`` and before transition
+from the corresponding node in script in runtime.
+
+Use-cases
+=========
+
+TODO
+
+See `tutorial about transition functions <../tutorials/tutorials.script.core.4_transitions>`__.
+
+Signature
+=========
+
+.. code-block:: python
+
+    def label(ctx: Context, pipeline: Pipeline) -> Tuple[str, str, float]:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is an instance of `NodeLabel3Type <../api/dff.script.core.types#NodeLabel3Type>`__,
+that is a tuple of target flow name (``str``), node name (``str``) and priority (``float``).
+
+Standard
+========
+
+There is a set of `standard label functions <../api/dff.script.conditions.std_labels>`__ defined.
+
+Exceptions
+==========
+
+If an exception occurs during this function execution, it will be reported during script validation stage
+(if any), otherwise it will be handled internally, only an exception message will be printed to ``stdout``.
+
+Responses
+~~~~~~~~~
+
+Description
+===========
+
+For some of the nodes in `Script <../api/dff.script.core.script#Script>`__ yielding constant response values
+might be not enough.
+For these cases each response can be represented as a Python function.
+
+These functions are executed on pipeline startup if ``validation_stage`` parameter of
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ is ``True`` and in the end
+of any node processing in runtime.
+
+Use-cases
+=========
+
+TODO
+
+See `tutorial about response functions <../tutorials/tutorials.script.core.3_responses>`__.
+
+Signature
+=========
+
+.. code-block:: python
+
+    def response(ctx: Context, pipeline: Pipeline) -> Message:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is an instance of `Message <../api/dff.script.core.message#Message>`__.
+
+Exceptions
+==========
+
+If an exception occurs during this function execution, it will be reported during script validation stage
+(if any), otherwise it will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the actor service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+
+Service functions
++++++++++++++++++
+
+Service handlers
+~~~~~~~~~~~~~~~~
+
+Description
+===========
+
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ services (other than `Actor <../api/dff.pipeline.pipeline.pipeline#ACTOR>`__)
+should be represented as functions.
+These functions can be run sequentially or combined into several asynchronous groups.
+The handlers can, for instance, process data, make web requests, read and write files, etc.
+
+The services are executed on every `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ run,
+they can run before or after `Actor <../api/dff.pipeline.pipeline.pipeline#ACTOR>`__ execution.
+
+Use-cases
+=========
+
+TODO
+
+See tutorials:
+
+#. `About synchronous services (basic) <../tutorials/tutorials.pipeline.3_pipeline_dict_with_services_basic>`__
+#. `About synchronous services (advanced) <../tutorials/tutorials.pipeline.3_pipeline_dict_with_services_full>`__
+#. `About asynchronous services (basic) <../tutorials/tutorials.pipeline.5_asynchronous_groups_and_services_basic>`__
+#. `About asynchronous services (advanced) <../tutorials/tutorials.pipeline.5_asynchronous_groups_and_services_full>`__
+
+Signatures
+==========
+
+.. code-block:: python
+
+    async def handler(ctx: Context) -> Any:
+        ...
+
+    async def handler(ctx: Context, pipeline: Pipeline) -> Any:
+        ...
+
+    async def handler(ctx: Context, pipeline: Pipeline, runtime_info: Dict) -> Any:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__,
+where ``runtime_info`` is a `runtime info dictionary <../api/dff.pipeline.types#ExtraHandlerRuntimeInfo>`__
+and the return value can be anything (it is not used).
+
+Exceptions
+==========
+
+If this function exceeds timeout (that implies that ``TimeoutError`` is thrown), it will be interrupted
+in parent `ServiceGroup <../api/dff.pipeline.service.group#ServiceGroup>`__ and an exception message will be printed to ``stdout``.
+If any other exception occurs during this function execution, it will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+
+Service conditions
+~~~~~~~~~~~~~~~~~~
+
+Description
+===========
+
+`Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ services (other than `Actor <../api/dff.pipeline.pipeline.pipeline#ACTOR>`__)
+can be executed conditionally.
+For that some special conditions should be used (that are in a way similar to `Script conditions and condition handlers`_).
+However, there is no such thing as ``condition handler`` function in pipeline.
+
+These conditions are only run before services they are related to, that can be any services **except for Actor**.
+
+Use-cases
+=========
+
+TODO
+
+See tutorials:
+
+#. `About service groups and conditions (basic) <../tutorials/tutorials.pipeline.4_groups_and_conditions_basic>`__
+#. `About service groups and conditions (advanced) <../tutorials/tutorials.pipeline.4_groups_and_conditions_full>`__
+
+Signature
+=========
+
+.. code-block:: python
+
+    def condition(ctx: Context, pipeline: Pipeline) -> bool:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and the return value is ``True`` if the service should be run and ``False`` otherwise.
+
+Standard
+========
+
+There is a set of `standard service condition functions <../api/dff.pipeline.conditions>`__ defined.
+
+Exceptions
+==========
+
+If any other exception occurs during this function execution, it will be handled on pipeline level,
+exception message will be printed to ``stdout`` and the service `state <../api/dff.pipeline.types#ComponentExecutionState>`__
+will be set to ``FAILED``.
+
+Extra handlers
+~~~~~~~~~~~~~~
+
+Description
+===========
+
+For any service in a `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__ special
+extra handler functions can be added.
+These functions can handle statistics collection, input data transformation
+or other pipeline functionality extension.
+
+These functions can be either added to `pipeline dict <../api/dff.pipeline.types#PipelineBuilder>`__
+or added to all services at once with `add_global_handler <../api/dff.pipeline.pipeline.pipeline#add_global_handler>`__
+function.
+The handlers can be executed before or after pipeline services.
+
+Use-cases
+=========
+
+TODO
+
+See tutorials:
+
+#. `About extra handlers (basic) <../tutorials/tutorials.pipeline.7_extra_handlers_basic>`__
+#. `About extra handlers (advanced) <../tutorials/tutorials.pipeline.7_extra_handlers_full>`__
+#. `About extra handlers and extensions <../tutorials/tutorials.pipeline.8_extra_handlers_and_extensions>`__
+
+Signatures
+==========
+
+.. code-block:: python
+
+    async def handler(ctx: Context) -> Any:
+        ...
+
+    async def handler(ctx: Context, pipeline: Pipeline) -> Any:
+        ...
+
+    async def handler(ctx: Context, pipeline: Pipeline, runtime_info: Dict) -> Any:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__,
+where ``runtime_info`` is a `runtime info dictionary <../api/dff.pipeline.types#ExtraHandlerRuntimeInfo>`__
+and the return value can be anything (it is not used).
+
+Exceptions
+==========
+
+If this function exceeds timeout (that implies that ``TimeoutError`` is thrown), it will be interrupted
+and an exception message will be printed to ``stdout``.
+If any other exception occurs during this function execution, it **will not** be handled on pipeline level,
+it will either be reported in parent `ServiceGroup <../api/dff.pipeline.service.group#ServiceGroup>`__ or interrupt pipeline execution.
+
+Statistics extractors
+~~~~~~~~~~~~~~~~~~~~~
+
+Description
+===========
+
+Statistics module introduces the concept of extractor functions.
+These are equivalent to pipeline's regular extra handlers (described above) except that
+they have to be decorated by Opentelemetry's `OtelInstrumentor <../api/dff.stats.instrumentor#OtelInstrumentor>`_
+class that proxies the output of these handlers and redirects it to Opentelemetry services.
+
+See the `extractor function tutorial <../tutorials/tutorials.stats.1_extractor_functions.py>`__ for reference.
+
+Use-cases
+=========
+
+TODO
+
+See `tutorial about extractor functions <../tutorials/tutorials.stats.1_extractor_functions>`__.
+
+Signature
+=========
+
+.. code-block:: python
+
+    def extractor(ctx: Context, pipeline: Pipeline, runtime_info: Dict) -> None:
+        ...
+
+where ``ctx`` is the current instance of `Context <../api/dff.script.core.context#Context>`__,
+where ``pipeline`` is the current instance of `Pipeline <../api/dff.pipeline.pipeline.pipeline#Pipeline>`__
+and ``runtime_info`` is a `runtime info dictionary <../api/dff.pipeline.types#ExtraHandlerRuntimeInfo>`__.
+
+Standard
+========
+
+There is a set of `standard statistics extractors <../api/dff.stats.default_extractors>`__ defined.
+
+Exceptions
+==========
+
+If an exception occurs during this function execution, it is not handled and will be thrown
+during `OtelInstrumentor <../api/dff.stats.instrumentor#OtelInstrumentor>`__ ``__call__``
+function execution.