chore: fixed flakiness of tests

ref: N25B-387

.env.example

@@ -0,0 +1,25 @@
+# Example .env file. To use, make a copy, call it ".env" (i.e. removing the ".example" suffix), then you edit values.
+# To make a variable apply, uncomment it (remove the "#" in front of the line).
+
+# First, some variables that are likely to be configured:
+
+# The hostname or IP address of the Control Backend.
+AGENT__CONTROL_BACKEND_HOST=localhost
+
+
+
+# Variables that are unlikely to be configured, you can probably ignore these:
+
+#AGENT__ACTUATION_RECEIVER_PORT=
+#AGENT__MAIN_RECEIVER_PORT=
+#AGENT__VIDEO_SENDER_PORT=
+#AGENT__AUDIO_SENDER_PORT=
+#VIDEO__CAMERA_INDEX=
+#VIDEO__RESOLUTION=
+#VIDEO__COLOR_SPACE=
+#VIDEO__FPS=
+#VIDEO__STREAM_NAME=
+#VIDEO__IMAGE_BUFFER=
+#AUDIO__SAMPLE_RATE=
+#AUDIO__CHUNK_SIZE=
+#AUDIO__CHANNELS=

.gitignore

@@ -220,3 +220,5 @@ __marimo__/
 # Docs
 docs/*
 !docs/conf.py
+!docs/installation/
+!docs/installation/**

README.md

@@ -8,90 +8,21 @@ This is an implementation for the Pepper robot, using the Pepper SDK and Python
 
 ## Installation
 
-### Linux (or WSL)
+- [Linux](./docs/installation/linux.md)
+- [macOS](./docs/installation/macos.md)
+- [Windows](./docs/installation/windows.md)
 
-Start off by installing [Pyenv](https://github.com/pyenv/pyenv?tab=readme-ov-file#installation) and walk through the steps outlined there (be sure to also add it to PATH). Also install the [Python build requirements](https://github.com/pyenv/pyenv/wiki#suggested-build-environment). Afterwards, install Python 2.7 and activate it for your current shell:
+
+
+### Git Hooks
+
+To activate automatic linting, formatting, branch name checks and commit message checks, run (after installing requirements):
 
 ```bash
-pyenv install 2.7
+pre-commit install
-pyenv shell 2.7
+pre-commit install --hook-type commit-msg
 ```
 
-You can check that this worked by typing
-
-```bash
-python -V
-```
-
-Which should return `Python 2.7.18`.
-
-Next, `cd` into this repository and create (and activate) a virtual environment:
-
-```bash
-cd <path to project>/
-python -m pip install virtualenv
-python -m virtualenv .venv
-source .venv/bin/activate
-```
-
-We depend on PortAudio for the `pyaudio` package, so install it with:
-
-```bash
-sudo apt install -y portaudio19-dev
-```
-
-On WSL, also install:
-
-```bash
-sudo apt install -y libasound2-plugins
-```
-
-Install the required packages with
-
-```bash
-pip install -r requirements.txt
-```
-
-Now we need to install the NaoQi SDK into our virtual environment, which we need to do manually. Begin by downloading the SDK:
-
-```bash
-wget https://community-static.aldebaran.com/resources/2.5.10/Python%20SDK/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
-```
-
-Next, move into the `site-packages` directory and extract the file you just downloaded:
-
-```bash
-cd .venv/lib/python2.7/site-packages/
-tar xvfz <path to SDK>/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
-rm <path to SDK>/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
-```
-
-Lastly, we need to inform our virtual environment where to find our newly installed package:
-
-```bash
-echo <path to project>/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.5.7.1-linux64/lib/python2.7/site-packages/ > pynaoqi-python2.7.pth
-```
-
-That's it! Verify that it works with
-
-```bash
-python -c "import qi; print(qi)"
-```
-
-You should now be able to run this project.
-
-### macOS
-
-Similar to Linux, but don't bother installing `pyenv` as it won't be able to install Python 2 on Apple Silicon. Instead, install Python 2.7.18 from the [Python website](https://www.python.org/downloads/release/python-2718/).
-
-Create the virtual environment as described above in the Linux section. Stop at the point where it shows you how to download the NaoQi SDK. Instead, use:
-
-```shell
-curl -OL https://community-static.aldebaran.com/resources/2.5.10/Python%20SDK/pynaoqi-python2.7-2.5.7.1-mac64.tar.gz
-```
-
-Then resume the steps from above.
-
 
 
 ## Usage
@@ -108,10 +39,15 @@ On Windows:
 $env:PYTHONPATH="src"; python -m robot_interface.main
 ```
 
-With both, if you want to connect to the actual robot (or simulator), pass the `--qi-url` argument.
+### Program Arguments
+
+If you want to connect to the actual robot (or simulator), pass the `--qi-url` argument.
 
 There's also a `--microphone` argument that can be used to choose a microphone to use. If not given, the program will try the default microphone. If you don't know the name of the microphone, pass the argument with any value, and it will list the names of available microphones.
 
+### Environment Variables
+
+You may use environment variables to change settings. Make a copy of the [`.env.example`](.env.example) file, name it `.env` and put it in the root directory. The file itself describes how to do the configuration.
 
 
 ## Testing
@@ -134,23 +70,6 @@ For coverage, add `--cov=robot_interface` as an argument to `pytest`.
 
 
 
-## Git Hooks
-
-To activate automatic linting, formatting, branch name checks and commit message checks, run (after installing requirements):
-
-```bash
-pre-commit install
-pre-commit install --hook-type commit-msg
-```
-
-You might get an error along the lines of `Can't install pre-commit with core.hooksPath` set. To fix this, simply unset the hooksPath by running:
-
-```bash
-git config --local --unset core.hooksPath
-```
-
-Then run the pre-commit install commands again.
-
 ## Documentation
 Generate documentation web pages using:
 

docs/installation/linux.md

@@ -0,0 +1,75 @@
+# Installation
+
+Of the Pepper Robot Interface on Linux (or WSL).
+
+Start off by installing [Pyenv](https://github.com/pyenv/pyenv?tab=readme-ov-file#installation) and walk through the steps outlined there (be sure to also add it to PATH). Also install the [Python build requirements](https://github.com/pyenv/pyenv/wiki#suggested-build-environment). Afterwards, install Python 2.7 and activate it for your current shell:
+
+```bash
+pyenv install 2.7
+pyenv shell 2.7
+```
+
+You can check that this worked by typing
+
+```bash
+python -V
+```
+
+Which should return `Python 2.7.18`.
+
+Next, `cd` into this repository and create (and activate) a virtual environment:
+
+```bash
+cd <path to project>/
+python -m pip install virtualenv
+python -m virtualenv .venv
+source .venv/bin/activate
+```
+
+We depend on PortAudio for the `pyaudio` package, so install it with:
+
+```bash
+sudo apt install -y portaudio19-dev
+```
+
+On WSL, also install:
+
+```bash
+sudo apt install -y libasound2-plugins
+```
+
+Install the required packages with
+
+```bash
+pip install -r requirements.txt
+```
+
+Now we need to install the NaoQi SDK into our virtual environment, which we need to do manually. Begin by downloading the SDK:
+
+```bash
+wget https://community-static.aldebaran.com/resources/2.5.10/Python%20SDK/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
+```
+
+Next, move into the `site-packages` directory and extract the file you just downloaded:
+
+```bash
+cd .venv/lib/python2.7/site-packages/
+tar xvfz <path to SDK>/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
+rm <path to SDK>/pynaoqi-python2.7-2.5.7.1-linux64.tar.gz
+```
+
+Lastly, we need to inform our virtual environment where to find our newly installed package:
+
+```bash
+echo <path to project>/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.5.7.1-linux64/lib/python2.7/site-packages/ > pynaoqi-python2.7.pth
+```
+
+That's it! Verify that it works with
+
+```bash
+python -c "import qi; print(qi)"
+```
+
+You should now be able to run this project.
+
+See the README for how to run.

docs/installation/macos.md

@@ -0,0 +1,106 @@
+# Installation
+
+Of the Pepper Robot Interface on macOS.
+
+
+
+## Python 2.7
+
+Install Python 2.7.18 from the [Python website](https://www.python.org/downloads/release/python-2718/).
+
+Check that it worked by executing
+
+```shell
+python2 -V
+```
+
+Which should return Python 2.7.18.
+
+
+
+## Virtual Environment
+
+Next, cd into this repository and create (and activate) a virtual environment:
+
+```shell
+cd /path/to/project/
+python2 -m pip install virtualenv
+python2 -m virtualenv .venv
+source .venv/bin/activate
+```
+
+We depend on PortAudio for the `pyaudio` package. If on Intel, run `brew install portaudio`. If on Apple Silicon, compile manually using the steps described in [the YouTrack article](https://utrechtuniversity.youtrack.cloud/articles/N25B-A-22/Install-PyAudio-for-Python-2-on-Apple-Silicon).
+
+Then install the required Python packages with
+
+```shell
+pip install -r requirements.txt
+```
+
+
+
+## NaoQi SDK
+
+We need to manually install the NaoQi SDK into our virtual environment. There are two options:
+
+1. Install a newer version (2.8) which will make running easier, but compatibility is uncertain.
+2. Install the version expected by the robot (2.5). This will complicate running slightly.
+
+### Option 1
+
+Download the SDK from [twirre.io](https://twirre.io/files/pynaoqi-python2.7-2.8.6.23-mac64-20191127_144231.tar.gz), or find one on the Aldebaran website, or an archived version on Web Archive.
+
+Extract it to `/path/to/project/.venv/lib/python2.7/site-packages/`.
+
+We need to inform our virtual environment where to find our newly installed package:
+
+```bash
+echo "/path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.8.6.23-mac64-20191127_144231/lib/python2.7/site-packages/" > /path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7.pth
+```
+
+Now continue with [verifying](#verifying).
+
+### Option 2
+
+This method of installation requires setting the `DYLD_LIBRARY_PATH` environment variable before running. How will be explained.
+
+Download the SDK from [twirre.io](https://twirre.io/files/pynaoqi-2.5.7.1-mac64-deps.tar.gz). This is a modified version of the one from Aldebaran, this one including required Choregraphe dependencies.
+
+Extract it to `/path/to/project/.venv/lib/python2.7/site-packages/`.
+
+We need to inform our virtual environment where to find our newly installed package:
+
+```shell
+echo "/path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.5.7.1-mac64/lib/python2.7/site-packages/" > /path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7.pth
+```
+
+Now, anytime before running you need to set the `DYLD_LIBRARY_PATH` environment variable.
+
+```shell
+export DYLD_LIBRARY_PATH="/path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.5.7.1-mac64/choregraphe_lib:${DYLD_LIBRARY_PATH}"
+```
+
+You may want to simplify environment activation with a script `activate.sh` like:
+
+```shell
+#!/bin/zsh
+
+export DYLD_LIBRARY_PATH="/path/to/project/.venv/lib/python2.7/site-packages/pynaoqi-python2.7-2.8.6.23-mac64-20191127_144231/choregraphe_lib:${DYLD_LIBRARY_PATH}"
+source .venv/bin/activate
+```
+
+[Verify](#verifying) if it works.
+
+
+
+## Verifying
+
+Verify that the NaoQI SDK installation works with
+
+```bash
+python -c "import qi; print(qi)"
+```
+
+If so, you should now be able to run this project.
+
+See the README for how to run.

docs/installation/windows.md

@@ -0,0 +1,44 @@
+# Installation
+
+Of the Pepper Robot Interface on Windows.
+
+Install Python 2.7.18 from [the Python website](https://www.python.org/downloads/release/python-2718/), choose the x86-64 installer (at the bottom of the page).
+
+To see if it worked:
+
+```shell
+py -2 -V
+```
+
+Which should return `Python 2.7.18`.
+
+Next, `cd` into this repository and create (and activate) a virtual environment:
+
+```bash
+cd <path to project>/
+py -2 -m pip install virtualenv
+py -2 -m virtualenv .venv
+.\.venv\Scripts\activate
+```
+
+Install the required packages with
+
+```bash
+pip install -r requirements.txt
+```
+
+Now we need to install the NaoQi SDK into our virtual environment, which we need to do manually. Download the SDK from [Aldebaran](https://community-static.aldebaran.com/resources/2.5.5/sdk-python/pynaoqi-python2.7-2.5.5.5-win32-vs2013.zip), [Web Archive](https://web.archive.org/web/20240120111043/https://community-static.aldebaran.com/resources/2.5.5/sdk-python/pynaoqi-python2.7-2.5.5.5-win32-vs2013.zip) or [twirre.io](https://twirre.io/files/pynaoqi-python2.7-2.8.6.23-win64-vs2015-20191127_152649.zip).
+
+Extract to `.\.venv\Lib\site-packages`.
+
+Create a file `.venv\Lib\site-packages\pynaoqi-python2.7.pth`, put the full path of `pynaoqi-python2.7-2.8.6.23-win64-vs2015-20191127_152649\lib\python2.7\Lib\site-packages` in there.
+
+Test if it worked by running:
+
+```bash
+python -c "import qi; print(qi)"
+```
+
+You should now be able to run this project.
+
+See the README for how to run.

requirements.txt

@@ -6,3 +6,4 @@ pytest-cov<3.0.0
 sphinx
 sphinx_rtd_theme
 pre-commit
+python-dotenv

src/robot_interface/core/config.py

@@ -1,95 +1,101 @@
 from __future__ import unicode_literals
 
+from robot_interface.utils.get_config import get_config
+
 
 class AgentSettings(object):
     """
     Agent port configuration.
 
-    :ivar actuating_receiver_port: Port for receiving actuation commands.
+    :ivar control_backend_host: Hostname of the control backend, defaults to "localhost".
-    :vartype actuating_receiver_port: int
+    :vartype control_backend_host: string
-    :ivar main_receiver_port: Port for receiving main messages.
+    :ivar actuation_receiver_port: Port for receiving actuation commands, defaults to 5557.
+    :vartype actuation_receiver_port: int
+    :ivar main_receiver_port: Port for receiving main messages, defaults to 5555.
     :vartype main_receiver_port: int
-    :ivar video_sender_port: Port used for sending video frames.
+    :ivar video_sender_port: Port used for sending video frames, defaults to 5556.
     :vartype video_sender_port: int
-    :ivar audio_sender_port: Port used for sending audio data.
+    :ivar audio_sender_port: Port used for sending audio data, defaults to 5558.
     :vartype audio_sender_port: int
     """
     def __init__(
-        self, 
+        self,
-        actuating_receiver_port=5557, 
+        control_backend_host=None,
-        main_receiver_port=5555,
+        actuation_receiver_port=None,
-        video_sender_port=5556, 
+        main_receiver_port=None,
-        audio_sender_port=5558,
+        video_sender_port=None,
+        audio_sender_port=None,
     ):
-        self.actuating_receiver_port = actuating_receiver_port
+        self.control_backend_host = get_config(control_backend_host, "AGENT__CONTROL_BACKEND_HOST", "localhost")
-        self.main_receiver_port = main_receiver_port
+        self.actuation_receiver_port = get_config(actuation_receiver_port, "AGENT__ACTUATION_RECEIVER_PORT", 5557, int)
-        self.video_sender_port = video_sender_port
+        self.main_receiver_port = get_config(main_receiver_port, "AGENT__MAIN_RECEIVER_PORT", 5555, int)
-        self.audio_sender_port = audio_sender_port
+        self.video_sender_port = get_config(video_sender_port, "AGENT__VIDEO_SENDER_PORT", 5556, int)
+        self.audio_sender_port = get_config(audio_sender_port, "AGENT__AUDIO_SENDER_PORT", 5558, int)
 
 
 class VideoConfig(object):
     """
     Video configuration constants.
 
-    :ivar camera_index: Index of the camera used.
+    :ivar camera_index: Index of the camera used, defaults to 0.
     :vartype camera_index: int
-    :ivar resolution: Video resolution mode.
+    :ivar resolution: Video resolution mode, defaults to 2.
     :vartype resolution: int
-    :ivar color_space: Color space identifier.
+    :ivar color_space: Color space identifier, defaults to 11.
     :vartype color_space: int
-    :ivar fps: Frames per second of the video stream.
+    :ivar fps: Frames per second of the video stream, defaults to 15.
     :vartype fps: int
-    :ivar stream_name: Name of the video stream.
+    :ivar stream_name: Name of the video stream, defaults to "Pepper Video".
     :vartype stream_name: str
-    :ivar image_buffer: Internal buffer size for video frames.
+    :ivar image_buffer: Internal buffer size for video frames, defaults to 6.
     :vartype image_buffer: int
     """
     def __init__(
         self,
-        camera_index=0,
+        camera_index=None,
-        resolution=2,
+        resolution=None,
-        color_space=11,
+        color_space=None,
-        fps=15,
+        fps=None,
-        stream_name="Pepper Video",
+        stream_name=None,
-        image_buffer=6,
+        image_buffer=None,
     ):
-        self.camera_index = camera_index
+        self.camera_index = get_config(camera_index, "VIDEO__CAMERA_INDEX", 0, int)
-        self.resolution = resolution
+        self.resolution = get_config(resolution, "VIDEO__RESOLUTION", 2, int)
-        self.color_space = color_space
+        self.color_space = get_config(color_space, "VIDEO__COLOR_SPACE", 11, int)
-        self.fps = fps
+        self.fps = get_config(fps, "VIDEO__FPS", 15, int)
-        self.stream_name = stream_name  
+        self.stream_name = get_config(stream_name, "VIDEO__STREAM_NAME", "Pepper Video")
-        self.image_buffer  = image_buffer
+        self.image_buffer = get_config(image_buffer, "VIDEO__IMAGE_BUFFER", 6, int)
 
 
 class AudioConfig(object):
     """
     Audio configuration constants.
 
-    :ivar sample_rate: Audio sampling rate in Hz.
+    :ivar sample_rate: Audio sampling rate in Hz, defaults to 16000.
     :vartype sample_rate: int
-    :ivar chunk_size: Size of audio chunks to capture/process.
+    :ivar chunk_size: Size of audio chunks to capture/process, defaults to 512.
     :vartype chunk_size: int
-    :ivar channels: Number of audio channels.
+    :ivar channels: Number of audio channels, defaults to 1.
     :vartype channels: int
     """
-    def __init__(self, sample_rate=16000, chunk_size=512, channels=1):
+    def __init__(self, sample_rate=None, chunk_size=None, channels=None):
-        self.sample_rate = sample_rate
+        self.sample_rate = get_config(sample_rate, "AUDIO__SAMPLE_RATE", 16000, int)
-        self.chunk_size = chunk_size
+        self.chunk_size = get_config(chunk_size, "AUDIO__CHUNK_SIZE", 512, int)
-        self.channels = channels
+        self.channels = get_config(channels, "AUDIO__CHANNELS", 1, int)
 
 
 class MainConfig(object):
     """
     Main system configuration.
 
-    :ivar poll_timeout_ms: Timeout for polling events, in milliseconds.
+    :ivar poll_timeout_ms: Timeout for polling events, in milliseconds, defaults to 100.
     :vartype poll_timeout_ms: int
-    :ivar max_handler_time_ms: Maximum allowed handler time, in milliseconds.
+    :ivar max_handler_time_ms: Maximum allowed handler time, in milliseconds, defaults to 50.
     :vartype max_handler_time_ms: int
     """
-    def __init__(self, poll_timeout_ms=100, max_handler_time_ms=50):
+    def __init__(self, poll_timeout_ms=None, max_handler_time_ms=None):
-        self.poll_timeout_ms = poll_timeout_ms
+        self.poll_timeout_ms = get_config(poll_timeout_ms, "MAIN__POLL_TIMEOUT_MS", 100, int)
-        self.max_handler_time_ms = max_handler_time_ms
+        self.max_handler_time_ms = get_config(max_handler_time_ms, "MAIN__MAX_HANDLER_TIME_MS", 50, int)
 
 
 class Settings(object):

src/robot_interface/endpoints/actuation_receiver.py

@@ -1,12 +1,16 @@
 from __future__ import unicode_literals  # So that we can log texts with Unicode characters
 import logging
+import time
+from threading import Thread
 
+import Queue
 import zmq
 
 from robot_interface.endpoints.receiver_base import ReceiverBase
 from robot_interface.state import state
 
 from robot_interface.core.config import settings
+from robot_interface.endpoints.gesture_settings import GestureTags
 
 
 class ActuationReceiver(ReceiverBase):
@@ -21,12 +25,22 @@ class ActuationReceiver(ReceiverBase):
 
     :ivar _tts_service: The text-to-speech service object from the Qi session.
     :vartype _tts_service: qi.Session | None
+
+    :ivar _animation_service: The animation/gesture service object from the Qi session.
+    :vartype _animation_service: qi.Session | None
     """
-    def __init__(self, zmq_context, port=settings.agent_settings.actuating_receiver_port):
+    def __init__(self, zmq_context, port=settings.agent_settings.actuation_receiver_port):
         super(ActuationReceiver, self).__init__("actuation")
         self.create_socket(zmq_context, zmq.SUB, port)
         self.socket.setsockopt_string(zmq.SUBSCRIBE, u"")  # Causes block if given in options
         self._tts_service = None
+        self._animation_service = None
+        self._message_queue = Queue.Queue()
+        self._gesture_queue = Queue.Queue()
+        self.message_thread = Thread(target=self._handle_messages)
+        self.message_thread.start()
+        self.gesture_thread = Thread(target=self._handle_gestures)
+        self.gesture_thread.start()
 
     def _handle_speech(self, message):
         """
@@ -53,8 +67,93 @@ class ActuationReceiver(ReceiverBase):
         if not self._tts_service:
             self._tts_service = state.qi_session.service("ALTextToSpeech")
 
-        # Returns instantly. Messages received while speaking will be queued.
+        if (message.get("is_priority")):
-        qi.async(self._tts_service.say, text)
+            # Bypass queue and speak immediately
+            self.clear_queue()
+            self._message_queue.put(text)
+            logging.debug("Force speaking immediately: {}".format(text))
+        else: 
+            self._message_queue.put(text)
+
+
+    def clear_queue(self):
+        """
+        Safely drains all pending messages from the queue.
+        """
+        logging.info("Message queue size: {}".format(self._message_queue.qsize()))
+        try:
+            while True:
+                # Remove items one by one without waiting
+                self._message_queue.get_nowait()
+        except Queue.Empty:
+            pass
+        logging.info("Message queue cleared.")
+
+    def clear_gesture_queue(self):
+        """
+        Safely drains all pending gestures from the gesture queue.
+        """
+        logging.info("Gesture queue size: {}".format(self._gesture_queue.qsize()))
+        try:
+            while True:
+                # Remove items one by one without waiting
+                self._gesture_queue.get_nowait()
+        except Queue.Empty:
+            pass
+        logging.info("Gesture queue cleared.")
+        logging.info("Gesture queue size: {}".format(self._gesture_queue.qsize()))
+
+
+    def _handle_gesture(self, message, is_single):
+        """
+        Handle a gesture actuation request.
+        
+        :param message: The gesture to do, must contain properties "endpoint" and "data".
+        :type message: dict
+
+        :param is_single: Whether it's a specific single gesture or a gesture tag.
+        :type is_single: bool
+        """
+
+        gesture = message.get("data")
+        if not gesture:
+            logging.warn("Received gesture to do, but it lacks data.")
+            return
+
+        if not isinstance(gesture, (str, unicode)):
+            logging.warn("Received gesture to do but it is not a string.")
+            return
+
+        logging.debug("Received gesture to do: {}".format(gesture))
+
+        if is_single:
+            if gesture not in GestureTags.single_gestures:
+                logging.warn("Received single gesture to do, but it does not exist in settings")
+                return
+        else:
+            if gesture not in GestureTags.tags:
+                logging.warn("Received single tag to do, but it does not exist in settings")
+                return
+
+        if not state.qi_session: return
+        # If state has a qi_session, we know that we can import qi
+        import qi  # Takes a while only the first time it's imported
+
+        if not self._animation_service:
+            self._animation_service = state.qi_session.service("ALAnimationPlayer")
+
+        # Play the gesture. Pepper comes with predefined animations like "Wave", "Greet", "Clap"
+        # You can also create custom animations using Choregraphe and upload them to the robot.
+        if (message.get("is_priority")):
+            # Clear queue and play
+            self.clear_gesture_queue()
+            logging.debug("Force playing gesture immediately: {}".format(gesture))
+        if is_single:
+            logging.debug("Adding single gesture to queue: {}".format(gesture))
+        else:
+            logging.debug("Adding tag gesture to queue: {}".format(gesture))
+        self._gesture_queue.put(gesture)
+
 
     def handle_message(self, message):
         """
@@ -65,3 +164,42 @@ class ActuationReceiver(ReceiverBase):
         """
         if message["endpoint"] == "actuate/speech":
             self._handle_speech(message)
+        if message["endpoint"] == "actuate/gesture/tag":
+            self._handle_gesture(message, False)
+        if message["endpoint"] == "actuate/gesture/single":
+            self._handle_gesture(message, True)
+
+    def _handle_messages(self):
+        while not state.exit_event.is_set():
+            try:
+                text = self._message_queue.get(timeout=0.1)
+                if not state.is_speaking: print("Started speaking.")
+                state.is_speaking = True
+                self._tts_service.say(text)
+            except Queue.Empty:
+                if state.is_speaking: print("Finished speaking.")
+                state.is_speaking = False
+            except RuntimeError:
+                logging.warn("Lost connection to Pepper. Please check if you're connected to the local WiFi and restart this application.")
+                state.exit_event.set() 
+    
+    def _handle_gestures(self):
+        while not state.exit_event.is_set():
+            try:
+                gesture = self._gesture_queue.get(timeout=0.1)
+                self._animation_service.run(gesture)
+            except Queue.Empty:
+                pass
+            except RuntimeError:
+                logging.warn("Lost connection to Pepper. Please check if you're connected to the local WiFi and restart this application.")
+                state.exit_event.set()       
+
+    def endpoint_description(self):
+        """
+        Extend the default endpoint description with gesture tags.
+        Returned during negotiate/ports so the CB knows available gestures.
+        """
+        desc = super(ActuationReceiver, self).endpoint_description()
+        desc["gestures"] = GestureTags.tags
+        desc["single_gestures"] = GestureTags.single_gestures
+        return desc

src/robot_interface/endpoints/gesture_settings.py

@@ -0,0 +1,412 @@
+class GestureTags:
+    tags = ["above", "affirmative", "afford", "agitated", "all", "allright", "alright", "any", 
+            "assuage", "assuage", "attemper", "back", "bashful", "beg", "beseech", "blank", 
+            "body language", "bored", "bow", "but", "call", "calm", "choose", "choice", "cloud", 
+            "cogitate", "cool", "crazy", "disappointed", "down", "earth", "empty", "embarrassed", 
+            "enthusiastic", "entire", "estimate", "except", "exalted", "excited", "explain", "far", 
+            "field", "floor", "forlorn", "friendly", "front", "frustrated", "gentle", "gift", 
+            "give", "ground", "happy", "hello", "her", "here", "hey", "hi", "him", "hopeless", 
+            "hysterical", "I", "implore", "indicate", "joyful", "me", "meditate", "modest", 
+            "negative", "nervous", "no", "not know", "nothing", "offer", "ok", "once upon a time", 
+            "oppose", "or", "pacify", "pick", "placate", "please", "present", "proffer", "quiet", 
+            "reason", "refute", "reject", "rousing", "sad", "select", "shamefaced", "show", 
+            "show sky", "sky", "soothe", "sun", "supplicate", "tablet", "tall", "them", "there", 
+            "think", "timid", "top", "unless", "up", "upstairs", "void", "warm", "winner", "yeah", 
+            "yes", "yoo-hoo", "you", "your", "zero", "zestful"]
+    
+    single_gestures = [
+        "animations/Stand/BodyTalk/Listening/Listening_1",
+        "animations/Stand/BodyTalk/Listening/Listening_2",
+        "animations/Stand/BodyTalk/Listening/Listening_3",
+        "animations/Stand/BodyTalk/Listening/Listening_4",
+        "animations/Stand/BodyTalk/Listening/Listening_5",
+        "animations/Stand/BodyTalk/Listening/Listening_6",
+        "animations/Stand/BodyTalk/Listening/Listening_7",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_1",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_10",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_11",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_12",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_13",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_14",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_15",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_16",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_2",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_3",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_4",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_5",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_6",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_7",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_8",
+        "animations/Stand/BodyTalk/Speaking/BodyTalk_9",
+        "animations/Stand/BodyTalk/Thinking/Remember_1",
+        "animations/Stand/BodyTalk/Thinking/Remember_2",
+        "animations/Stand/BodyTalk/Thinking/Remember_3",
+        "animations/Stand/BodyTalk/Thinking/ThinkingLoop_1",
+        "animations/Stand/BodyTalk/Thinking/ThinkingLoop_2",
+        "animations/Stand/Emotions/Negative/Angry_1",
+        "animations/Stand/Emotions/Negative/Angry_2",
+        "animations/Stand/Emotions/Negative/Angry_3",
+        "animations/Stand/Emotions/Negative/Angry_4",
+        "animations/Stand/Emotions/Negative/Anxious_1",
+        "animations/Stand/Emotions/Negative/Bored_1",
+        "animations/Stand/Emotions/Negative/Bored_2",
+        "animations/Stand/Emotions/Negative/Disappointed_1",
+        "animations/Stand/Emotions/Negative/Exhausted_1",
+        "animations/Stand/Emotions/Negative/Exhausted_2",
+        "animations/Stand/Emotions/Negative/Fear_1",
+        "animations/Stand/Emotions/Negative/Fear_2",
+        "animations/Stand/Emotions/Negative/Fearful_1",
+        "animations/Stand/Emotions/Negative/Frustrated_1",
+        "animations/Stand/Emotions/Negative/Humiliated_1",
+        "animations/Stand/Emotions/Negative/Hurt_1",
+        "animations/Stand/Emotions/Negative/Hurt_2",
+        "animations/Stand/Emotions/Negative/Late_1",
+        "animations/Stand/Emotions/Negative/Sad_1",
+        "animations/Stand/Emotions/Negative/Sad_2",
+        "animations/Stand/Emotions/Negative/Shocked_1",
+        "animations/Stand/Emotions/Negative/Sorry_1",
+        "animations/Stand/Emotions/Negative/Surprise_1",
+        "animations/Stand/Emotions/Negative/Surprise_2",
+        "animations/Stand/Emotions/Negative/Surprise_3",
+        "animations/Stand/Emotions/Neutral/Alienated_1",
+        "animations/Stand/Emotions/Neutral/AskForAttention_1",
+        "animations/Stand/Emotions/Neutral/AskForAttention_2",
+        "animations/Stand/Emotions/Neutral/AskForAttention_3",
+        "animations/Stand/Emotions/Neutral/Cautious_1",
+        "animations/Stand/Emotions/Neutral/Confused_1",
+        "animations/Stand/Emotions/Neutral/Determined_1",
+        "animations/Stand/Emotions/Neutral/Embarrassed_1",
+        "animations/Stand/Emotions/Neutral/Hesitation_1",
+        "animations/Stand/Emotions/Neutral/Innocent_1",
+        "animations/Stand/Emotions/Neutral/Lonely_1",
+        "animations/Stand/Emotions/Neutral/Mischievous_1",
+        "animations/Stand/Emotions/Neutral/Puzzled_1",
+        "animations/Stand/Emotions/Neutral/Sneeze",
+        "animations/Stand/Emotions/Neutral/Stubborn_1",
+        "animations/Stand/Emotions/Neutral/Suspicious_1",
+        "animations/Stand/Emotions/Positive/Amused_1",
+        "animations/Stand/Emotions/Positive/Confident_1",
+        "animations/Stand/Emotions/Positive/Ecstatic_1",
+        "animations/Stand/Emotions/Positive/Enthusiastic_1",
+        "animations/Stand/Emotions/Positive/Excited_1",
+        "animations/Stand/Emotions/Positive/Excited_2",
+        "animations/Stand/Emotions/Positive/Excited_3",
+        "animations/Stand/Emotions/Positive/Happy_1",
+        "animations/Stand/Emotions/Positive/Happy_2",
+        "animations/Stand/Emotions/Positive/Happy_3",
+        "animations/Stand/Emotions/Positive/Happy_4",
+        "animations/Stand/Emotions/Positive/Hungry_1",
+        "animations/Stand/Emotions/Positive/Hysterical_1",
+        "animations/Stand/Emotions/Positive/Interested_1",
+        "animations/Stand/Emotions/Positive/Interested_2",
+        "animations/Stand/Emotions/Positive/Laugh_1",
+        "animations/Stand/Emotions/Positive/Laugh_2",
+        "animations/Stand/Emotions/Positive/Laugh_3",
+        "animations/Stand/Emotions/Positive/Mocker_1",
+        "animations/Stand/Emotions/Positive/Optimistic_1",
+        "animations/Stand/Emotions/Positive/Peaceful_1",
+        "animations/Stand/Emotions/Positive/Proud_1",
+        "animations/Stand/Emotions/Positive/Proud_2",
+        "animations/Stand/Emotions/Positive/Proud_3",
+        "animations/Stand/Emotions/Positive/Relieved_1",
+        "animations/Stand/Emotions/Positive/Shy_1",
+        "animations/Stand/Emotions/Positive/Shy_2",
+        "animations/Stand/Emotions/Positive/Sure_1",
+        "animations/Stand/Emotions/Positive/Winner_1",
+        "animations/Stand/Emotions/Positive/Winner_2",
+        "animations/Stand/Gestures/Angry_1",
+        "animations/Stand/Gestures/Angry_2",
+        "animations/Stand/Gestures/Angry_3",
+        "animations/Stand/Gestures/BowShort_1",
+        "animations/Stand/Gestures/BowShort_2",
+        "animations/Stand/Gestures/BowShort_3",
+        "animations/Stand/Gestures/But_1",
+        "animations/Stand/Gestures/CalmDown_1",
+        "animations/Stand/Gestures/CalmDown_2",
+        "animations/Stand/Gestures/CalmDown_3",
+        "animations/Stand/Gestures/CalmDown_4",
+        "animations/Stand/Gestures/CalmDown_5",
+        "animations/Stand/Gestures/CalmDown_6",
+        "animations/Stand/Gestures/Choice_1",
+        "animations/Stand/Gestures/ComeOn_1",
+        "animations/Stand/Gestures/Confused_1",
+        "animations/Stand/Gestures/Confused_2",
+        "animations/Stand/Gestures/CountFive_1",
+        "animations/Stand/Gestures/CountFour_1",
+        "animations/Stand/Gestures/CountMore_1",
+        "animations/Stand/Gestures/CountOne_1",
+        "animations/Stand/Gestures/CountThree_1",
+        "animations/Stand/Gestures/CountTwo_1",
+        "animations/Stand/Gestures/Desperate_1",
+        "animations/Stand/Gestures/Desperate_2",
+        "animations/Stand/Gestures/Desperate_3",
+        "animations/Stand/Gestures/Desperate_4",
+        "animations/Stand/Gestures/Desperate_5",
+        "animations/Stand/Gestures/DontUnderstand_1",
+        "animations/Stand/Gestures/Enthusiastic_3",
+        "animations/Stand/Gestures/Enthusiastic_4",
+        "animations/Stand/Gestures/Enthusiastic_5",
+        "animations/Stand/Gestures/Everything_1",
+        "animations/Stand/Gestures/Everything_2",
+        "animations/Stand/Gestures/Everything_3",
+        "animations/Stand/Gestures/Everything_4",
+        "animations/Stand/Gestures/Everything_6",
+        "animations/Stand/Gestures/Excited_1",
+        "animations/Stand/Gestures/Explain_1",
+        "animations/Stand/Gestures/Explain_10",
+        "animations/Stand/Gestures/Explain_11",
+        "animations/Stand/Gestures/Explain_2",
+        "animations/Stand/Gestures/Explain_3",
+        "animations/Stand/Gestures/Explain_4",
+        "animations/Stand/Gestures/Explain_5",
+        "animations/Stand/Gestures/Explain_6",
+        "animations/Stand/Gestures/Explain_7",
+        "animations/Stand/Gestures/Explain_8",
+        "animations/Stand/Gestures/Far_1",
+        "animations/Stand/Gestures/Far_2",
+        "animations/Stand/Gestures/Far_3",
+        "animations/Stand/Gestures/Follow_1",
+        "animations/Stand/Gestures/Give_1",
+        "animations/Stand/Gestures/Give_2",
+        "animations/Stand/Gestures/Give_3",
+        "animations/Stand/Gestures/Give_4",
+        "animations/Stand/Gestures/Give_5",
+        "animations/Stand/Gestures/Give_6",
+        "animations/Stand/Gestures/Great_1",
+        "animations/Stand/Gestures/HeSays_1",
+        "animations/Stand/Gestures/HeSays_2",
+        "animations/Stand/Gestures/HeSays_3",
+        "animations/Stand/Gestures/Hey_1",
+        "animations/Stand/Gestures/Hey_10",
+        "animations/Stand/Gestures/Hey_2",
+        "animations/Stand/Gestures/Hey_3",
+        "animations/Stand/Gestures/Hey_4",
+        "animations/Stand/Gestures/Hey_6",
+        "animations/Stand/Gestures/Hey_7",
+        "animations/Stand/Gestures/Hey_8",
+        "animations/Stand/Gestures/Hey_9",
+        "animations/Stand/Gestures/Hide_1",
+        "animations/Stand/Gestures/Hot_1",
+        "animations/Stand/Gestures/Hot_2",
+        "animations/Stand/Gestures/IDontKnow_1",
+        "animations/Stand/Gestures/IDontKnow_2",
+        "animations/Stand/Gestures/IDontKnow_3",
+        "animations/Stand/Gestures/IDontKnow_4",
+        "animations/Stand/Gestures/IDontKnow_5",
+        "animations/Stand/Gestures/IDontKnow_6",
+        "animations/Stand/Gestures/Joy_1",
+        "animations/Stand/Gestures/Kisses_1",
+        "animations/Stand/Gestures/Look_1",
+        "animations/Stand/Gestures/Look_2",
+        "animations/Stand/Gestures/Maybe_1",
+        "animations/Stand/Gestures/Me_1",
+        "animations/Stand/Gestures/Me_2",
+        "animations/Stand/Gestures/Me_4",
+        "animations/Stand/Gestures/Me_7",
+        "animations/Stand/Gestures/Me_8",
+        "animations/Stand/Gestures/Mime_1",
+        "animations/Stand/Gestures/Mime_2",
+        "animations/Stand/Gestures/Next_1",
+        "animations/Stand/Gestures/No_1",
+        "animations/Stand/Gestures/No_2",
+        "animations/Stand/Gestures/No_3",
+        "animations/Stand/Gestures/No_4",
+        "animations/Stand/Gestures/No_5",
+        "animations/Stand/Gestures/No_6",
+        "animations/Stand/Gestures/No_7",
+        "animations/Stand/Gestures/No_8",
+        "animations/Stand/Gestures/No_9",
+        "animations/Stand/Gestures/Nothing_1",
+        "animations/Stand/Gestures/Nothing_2",
+        "animations/Stand/Gestures/OnTheEvening_1",
+        "animations/Stand/Gestures/OnTheEvening_2",
+        "animations/Stand/Gestures/OnTheEvening_3",
+        "animations/Stand/Gestures/OnTheEvening_4",
+        "animations/Stand/Gestures/OnTheEvening_5",
+        "animations/Stand/Gestures/Please_1",
+        "animations/Stand/Gestures/Please_2",
+        "animations/Stand/Gestures/Please_3",
+        "animations/Stand/Gestures/Reject_1",
+        "animations/Stand/Gestures/Reject_2",
+        "animations/Stand/Gestures/Reject_3",
+        "animations/Stand/Gestures/Reject_4",
+        "animations/Stand/Gestures/Reject_5",
+        "animations/Stand/Gestures/Reject_6",
+        "animations/Stand/Gestures/Salute_1",
+        "animations/Stand/Gestures/Salute_2",
+        "animations/Stand/Gestures/Salute_3",
+        "animations/Stand/Gestures/ShowFloor_1",
+        "animations/Stand/Gestures/ShowFloor_2",
+        "animations/Stand/Gestures/ShowFloor_3",
+        "animations/Stand/Gestures/ShowFloor_4",
+        "animations/Stand/Gestures/ShowFloor_5",
+        "animations/Stand/Gestures/ShowSky_1",
+        "animations/Stand/Gestures/ShowSky_10",
+        "animations/Stand/Gestures/ShowSky_11",
+        "animations/Stand/Gestures/ShowSky_12",
+        "animations/Stand/Gestures/ShowSky_2",
+        "animations/Stand/Gestures/ShowSky_3",
+        "animations/Stand/Gestures/ShowSky_4",
+        "animations/Stand/Gestures/ShowSky_5",
+        "animations/Stand/Gestures/ShowSky_6",
+        "animations/Stand/Gestures/ShowSky_7",
+        "animations/Stand/Gestures/ShowSky_8",
+        "animations/Stand/Gestures/ShowSky_9",
+        "animations/Stand/Gestures/ShowTablet_1",
+        "animations/Stand/Gestures/ShowTablet_2",
+        "animations/Stand/Gestures/ShowTablet_3",
+        "animations/Stand/Gestures/Shy_1",
+        "animations/Stand/Gestures/Stretch_1",
+        "animations/Stand/Gestures/Stretch_2",
+        "animations/Stand/Gestures/Surprised_1",
+        "animations/Stand/Gestures/TakePlace_1",
+        "animations/Stand/Gestures/TakePlace_2",
+        "animations/Stand/Gestures/Take_1",
+        "animations/Stand/Gestures/Thinking_1",
+        "animations/Stand/Gestures/Thinking_2",
+        "animations/Stand/Gestures/Thinking_3",
+        "animations/Stand/Gestures/Thinking_4",
+        "animations/Stand/Gestures/Thinking_5",
+        "animations/Stand/Gestures/Thinking_6",
+        "animations/Stand/Gestures/Thinking_7",
+        "animations/Stand/Gestures/Thinking_8",
+        "animations/Stand/Gestures/This_1",
+        "animations/Stand/Gestures/This_10",
+        "animations/Stand/Gestures/This_11",
+        "animations/Stand/Gestures/This_12",
+        "animations/Stand/Gestures/This_13",
+        "animations/Stand/Gestures/This_14",
+        "animations/Stand/Gestures/This_15",
+        "animations/Stand/Gestures/This_2",
+        "animations/Stand/Gestures/This_3",
+        "animations/Stand/Gestures/This_4",
+        "animations/Stand/Gestures/This_5",
+        "animations/Stand/Gestures/This_6",
+        "animations/Stand/Gestures/This_7",
+        "animations/Stand/Gestures/This_8",
+        "animations/Stand/Gestures/This_9",
+        "animations/Stand/Gestures/WhatSThis_1",
+        "animations/Stand/Gestures/WhatSThis_10",
+        "animations/Stand/Gestures/WhatSThis_11",
+        "animations/Stand/Gestures/WhatSThis_12",
+        "animations/Stand/Gestures/WhatSThis_13",
+        "animations/Stand/Gestures/WhatSThis_14",
+        "animations/Stand/Gestures/WhatSThis_15",
+        "animations/Stand/Gestures/WhatSThis_16",
+        "animations/Stand/Gestures/WhatSThis_2",
+        "animations/Stand/Gestures/WhatSThis_3",
+        "animations/Stand/Gestures/WhatSThis_4",
+        "animations/Stand/Gestures/WhatSThis_5",
+        "animations/Stand/Gestures/WhatSThis_6",
+        "animations/Stand/Gestures/WhatSThis_7",
+        "animations/Stand/Gestures/WhatSThis_8",
+        "animations/Stand/Gestures/WhatSThis_9",
+        "animations/Stand/Gestures/Whisper_1",
+        "animations/Stand/Gestures/Wings_1",
+        "animations/Stand/Gestures/Wings_2",
+        "animations/Stand/Gestures/Wings_3",
+        "animations/Stand/Gestures/Wings_4",
+        "animations/Stand/Gestures/Wings_5",
+        "animations/Stand/Gestures/Yes_1",
+        "animations/Stand/Gestures/Yes_2",
+        "animations/Stand/Gestures/Yes_3",
+        "animations/Stand/Gestures/YouKnowWhat_1",
+        "animations/Stand/Gestures/YouKnowWhat_2",
+        "animations/Stand/Gestures/YouKnowWhat_3",
+        "animations/Stand/Gestures/YouKnowWhat_4",
+        "animations/Stand/Gestures/YouKnowWhat_5",
+        "animations/Stand/Gestures/YouKnowWhat_6",
+        "animations/Stand/Gestures/You_1",
+        "animations/Stand/Gestures/You_2",
+        "animations/Stand/Gestures/You_3",
+        "animations/Stand/Gestures/You_4",
+        "animations/Stand/Gestures/You_5",
+        "animations/Stand/Gestures/Yum_1",
+        "animations/Stand/Reactions/EthernetOff_1",
+        "animations/Stand/Reactions/EthernetOn_1",
+        "animations/Stand/Reactions/Heat_1",
+        "animations/Stand/Reactions/Heat_2",
+        "animations/Stand/Reactions/LightShine_1",
+        "animations/Stand/Reactions/LightShine_2",
+        "animations/Stand/Reactions/LightShine_3",
+        "animations/Stand/Reactions/LightShine_4",
+        "animations/Stand/Reactions/SeeColor_1",
+        "animations/Stand/Reactions/SeeColor_2",
+        "animations/Stand/Reactions/SeeColor_3",
+        "animations/Stand/Reactions/SeeSomething_1",
+        "animations/Stand/Reactions/SeeSomething_3",
+        "animations/Stand/Reactions/SeeSomething_4",
+        "animations/Stand/Reactions/SeeSomething_5",
+        "animations/Stand/Reactions/SeeSomething_6",
+        "animations/Stand/Reactions/SeeSomething_7",
+        "animations/Stand/Reactions/SeeSomething_8",
+        "animations/Stand/Reactions/ShakeBody_1",
+        "animations/Stand/Reactions/ShakeBody_2",
+        "animations/Stand/Reactions/ShakeBody_3",
+        "animations/Stand/Reactions/TouchHead_1",
+        "animations/Stand/Reactions/TouchHead_2",
+        "animations/Stand/Reactions/TouchHead_3",
+        "animations/Stand/Reactions/TouchHead_4",
+        "animations/Stand/Waiting/AirGuitar_1",
+        "animations/Stand/Waiting/BackRubs_1",
+        "animations/Stand/Waiting/Bandmaster_1",
+        "animations/Stand/Waiting/Binoculars_1",
+        "animations/Stand/Waiting/BreathLoop_1",
+        "animations/Stand/Waiting/BreathLoop_2",
+        "animations/Stand/Waiting/BreathLoop_3",
+        "animations/Stand/Waiting/CallSomeone_1",
+        "animations/Stand/Waiting/Drink_1",
+        "animations/Stand/Waiting/DriveCar_1",
+        "animations/Stand/Waiting/Fitness_1",
+        "animations/Stand/Waiting/Fitness_2",
+        "animations/Stand/Waiting/Fitness_3",
+        "animations/Stand/Waiting/FunnyDancer_1",
+        "animations/Stand/Waiting/HappyBirthday_1",
+        "animations/Stand/Waiting/Helicopter_1",
+        "animations/Stand/Waiting/HideEyes_1",
+        "animations/Stand/Waiting/HideHands_1",
+        "animations/Stand/Waiting/Innocent_1",
+        "animations/Stand/Waiting/Knight_1",
+        "animations/Stand/Waiting/KnockEye_1",
+        "animations/Stand/Waiting/KungFu_1",
+        "animations/Stand/Waiting/LookHand_1",
+        "animations/Stand/Waiting/LookHand_2",
+        "animations/Stand/Waiting/LoveYou_1",
+        "animations/Stand/Waiting/Monster_1",
+        "animations/Stand/Waiting/MysticalPower_1",
+        "animations/Stand/Waiting/PlayHands_1",
+        "animations/Stand/Waiting/PlayHands_2",
+        "animations/Stand/Waiting/PlayHands_3",
+        "animations/Stand/Waiting/Relaxation_1",
+        "animations/Stand/Waiting/Relaxation_2",
+        "animations/Stand/Waiting/Relaxation_3",
+        "animations/Stand/Waiting/Relaxation_4",
+        "animations/Stand/Waiting/Rest_1",
+        "animations/Stand/Waiting/Robot_1",
+        "animations/Stand/Waiting/ScratchBack_1",
+        "animations/Stand/Waiting/ScratchBottom_1",
+        "animations/Stand/Waiting/ScratchEye_1",
+        "animations/Stand/Waiting/ScratchHand_1",
+        "animations/Stand/Waiting/ScratchHead_1",
+        "animations/Stand/Waiting/ScratchLeg_1",
+        "animations/Stand/Waiting/ScratchTorso_1",
+        "animations/Stand/Waiting/ShowMuscles_1",
+        "animations/Stand/Waiting/ShowMuscles_2",
+        "animations/Stand/Waiting/ShowMuscles_3",
+        "animations/Stand/Waiting/ShowMuscles_4",
+        "animations/Stand/Waiting/ShowMuscles_5",
+        "animations/Stand/Waiting/ShowSky_1",
+        "animations/Stand/Waiting/ShowSky_2",
+        "animations/Stand/Waiting/SpaceShuttle_1",
+        "animations/Stand/Waiting/Stretch_1",
+        "animations/Stand/Waiting/Stretch_2",
+        "animations/Stand/Waiting/TakePicture_1",
+        "animations/Stand/Waiting/Taxi_1",
+        "animations/Stand/Waiting/Think_1",
+        "animations/Stand/Waiting/Think_2",
+        "animations/Stand/Waiting/Think_3",
+        "animations/Stand/Waiting/Think_4",
+        "animations/Stand/Waiting/Waddle_1",
+        "animations/Stand/Waiting/Waddle_2",
+        "animations/Stand/Waiting/WakeUp_1",
+        "animations/Stand/Waiting/Zombie_1"]         

src/robot_interface/endpoints/main_receiver.py

@@ -5,6 +5,7 @@ from robot_interface.state import state
 
 from robot_interface.core.config import settings
 
+
 class MainReceiver(ReceiverBase):
     """
     The main receiver endpoint, responsible for handling ping and negotiation requests.
@@ -12,10 +13,12 @@ class MainReceiver(ReceiverBase):
     :param zmq_context: The ZeroMQ context to use.
     :type zmq_context: zmq.Context
 
-    :param port: The port to use.
+    :param port: The port to use, defaults to value in `settings.agent_settings.main_receiver_port`.
     :type port: int
     """
-    def __init__(self, zmq_context, port=settings.agent_settings.main_receiver_port):
+    def __init__(self, zmq_context, port=None):
+        if port is None:
+            port = settings.agent_settings.main_receiver_port
         super(MainReceiver, self).__init__("main")
         self.create_socket(zmq_context, zmq.REP, port, bind=False)
 

src/robot_interface/endpoints/socket_base.py

@@ -2,6 +2,8 @@ from abc import ABCMeta
 
 import zmq
 
+from robot_interface.core.config import settings
+
 
 class SocketBase(object):
     """
@@ -59,7 +61,7 @@ class SocketBase(object):
         if bind:
             self.socket.bind("tcp://*:{}".format(port))
         else:
-            self.socket.connect("tcp://localhost:{}".format(port))
+            self.socket.connect("tcp://{}:{}".format(settings.agent_settings.control_backend_host, port))
 
     def close(self):
         """Close the ZeroMQ socket."""

src/robot_interface/state.py

@@ -30,6 +30,7 @@ class State(object):
         self.exit_event = None
         self.sockets = []  
         self.qi_session = None 
+        self.is_speaking = False
 
     def initialize(self):
         """

src/robot_interface/utils/get_config.py

@@ -0,0 +1,32 @@
+import os
+from dotenv import load_dotenv
+
+load_dotenv()
+
+
+def get_config(value, env, default, cast=None):
+    """
+    Small utility to get a configuration value, returns `value` if it is not None, else it will try to get the
+    environment variable cast with `cast`. If the environment variable is not set, it will return `default`.
+
+    :param value: The value to check.
+    :type value: Any
+    :param env: The environment variable to check.
+    :type env: string
+    :param default: The default value to return if the environment variable is not set.
+    :type default: Any
+    :param cast: A function to use to cast the environment variable. Must support string input.
+    :type cast: Callable[[Any], Any], optional
+
+    :return: The value, the environment variable value, or the default.
+    :rtype: Any
+    """
+    if value is not None:
+        return value
+
+    env = os.environ.get(env, default)
+
+    if cast is None:
+        return env
+
+    return cast(env)

test/conftest.py

@@ -0,0 +1,10 @@
+from mock import patch, MagicMock
+
+import pytest
+
+
+@pytest.fixture(autouse=True)
+def mock_zmq_context():
+    with patch("zmq.Context") as mock:
+        mock.instance.return_value = MagicMock()
+        yield mock

test/integration/test_config.py

@@ -0,0 +1,32 @@
+from mock import patch, mock
+
+from robot_interface.core.config import Settings
+from robot_interface.endpoints.main_receiver import MainReceiver
+
+
+def test_environment_variables(monkeypatch):
+    """
+    When environment variables are set, creating settings should use these.
+    """
+    monkeypatch.setenv("AGENT__CONTROL_BACKEND_HOST", "some_value_that_should_be_different")
+
+    settings = Settings()
+
+    assert settings.agent_settings.control_backend_host == "some_value_that_should_be_different"
+
+
+@patch("robot_interface.endpoints.main_receiver.settings")
+@patch("robot_interface.endpoints.socket_base.settings")
+def test_create_endpoint_custom_host(base_settings, main_settings):
+    """
+    When a custom host is given in the settings, check that an endpoint's socket connects to it.
+    """
+    fake_context = mock.Mock()
+    fake_socket = mock.Mock()
+    fake_context.socket.return_value = fake_socket
+    base_settings.agent_settings.control_backend_host = "not_localhost"
+    main_settings.agent_settings.main_receiver_port = 9999
+
+    _ = MainReceiver(fake_context)
+
+    fake_socket.connect.assert_called_once_with("tcp://not_localhost:9999")

test/unit/test_actuation_receiver.py

@@ -1,64 +1,114 @@
 import sys
 
+import time
 import mock
 import pytest
 import zmq
-
+import Queue
 from robot_interface.endpoints.actuation_receiver import ActuationReceiver
+from robot_interface.endpoints.gesture_settings import GestureTags
 
-
+def test_force_speech_clears_queue(mocker):
-@pytest.fixture
-def zmq_context():
     """
-    A pytest fixture that creates and yields a ZMQ context.
+    Tests that a force speech message clears the existing queue
-
+    and places the high-priority message at the front.
-    :return: An initialized ZeroMQ context.
-    :rtype: zmq.Context
     """
-    context = zmq.Context()
+    mocker.patch("threading.Thread")
-    yield context
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi 
 
+    mock_tts_service = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_tts_service
 
-def test_handle_unimplemented_endpoint(zmq_context):
+    # Use Mock Context
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    receiver._message_queue.put("old_message_1")
+    receiver._message_queue.put("old_message_2")
+    
+    assert receiver._message_queue.qsize() == 2
+
+    force_msg = {
+        "endpoint": "actuate/speech", 
+        "data": "Emergency Notification",
+        "is_priority": True,
+    }
+    receiver.handle_message(force_msg)
+
+    assert receiver._message_queue.qsize() == 1
+    queued_item = receiver._message_queue.get()
+    assert queued_item == "Emergency Notification"
+
+def test_handle_unimplemented_endpoint(mocker):
     """
-    Tests that the ``ActuationReceiver.handle_message`` method can
+    Tests handling of unknown endpoints.
-    handle an unknown or unimplemented endpoint without raising an error.
     """
-    receiver = ActuationReceiver(zmq_context)
+    mocker.patch("threading.Thread")
-    # Should not error
+    
+    # Use Mock Context
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
     receiver.handle_message({
         "endpoint": "some_endpoint_that_definitely_does_not_exist",
         "data": None,
     })
 
-
+def test_speech_message_no_data(mocker):
-def test_speech_message_no_data(zmq_context, mocker):
     """
-    Tests that the message handler logs a warning when a speech actuation
+    Tests that if the message data is empty, the receiver returns immediately
-    request (`actuate/speech`) is received but contains empty string data.
+    WITHOUT attempting to access the global robot state or session.
     """
-    mock_warn = mocker.patch("logging.warn")
+    # 1. Prevent background threads from running
+    mocker.patch("threading.Thread")
+    
+    # 2. Mock the global state object
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
 
-    receiver = ActuationReceiver(zmq_context)
+    # 3. Create a PropertyMock to track whenever 'qi_session' is accessed
+    # We attach it to the class type of the mock so it acts like a real property
+    mock_session_prop = mock.PropertyMock(return_value=None)
+    type(mock_state).qi_session = mock_session_prop
+
+    # 4. Initialize Receiver (Mocking the context to avoid ZMQ errors)
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # 5. Send empty data
     receiver.handle_message({"endpoint": "actuate/speech", "data": ""})
 
-    mock_warn.assert_called_with(mock.ANY)
+    # 6. Assertion: 
+    # Because the code does `if not text: return` BEFORE `if not state.qi_session`,
+    # the state property should NEVER be read.
+    mock_session_prop.assert_not_called()
 
 
-def test_speech_message_invalid_data(zmq_context, mocker):
+def test_speech_message_invalid_data(mocker):
     """
-    Tests that the message handler logs a warning when a speech actuation
+    Tests that if the message data is not a string, the function returns.     
-    request (`actuate/speech`) is received with data that is not a string (e.g., a boolean).
+    :param mocker: Description
     """
-    mock_warn = mocker.patch("logging.warn")
+    mocker.patch("threading.Thread")
 
-    receiver = ActuationReceiver(zmq_context)
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+
+    mock_session_prop = mock.PropertyMock(return_value=None)
+    type(mock_state).qi_session = mock_session_prop
+
+    # Use Mock Context
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
     receiver.handle_message({"endpoint": "actuate/speech", "data": True})
 
-    mock_warn.assert_called_with(mock.ANY)
+    # Because the code does `if not text: return` BEFORE `if not state.qi_session`,
+    # the state property should NEVER be read.
+    mock_session_prop.assert_not_called()
 
-
+def test_speech_no_qi(mocker):
-def test_speech_no_qi(zmq_context, mocker):
     """
     Tests the actuation receiver's behavior when processing a speech request
     but the global state does not have an active QI session.
@@ -68,16 +118,21 @@ def test_speech_no_qi(zmq_context, mocker):
     mock_qi_session = mock.PropertyMock(return_value=None)
     type(mock_state).qi_session = mock_qi_session
 
-    receiver = ActuationReceiver(zmq_context)
+    mock_tts_service = mock.Mock()
+
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    receiver._tts_service = mock_tts_service
+
     receiver._handle_speech({"endpoint": "actuate/speech", "data": "Some message to speak."})
 
-    mock_qi_session.assert_called()
+    receiver._tts_service.assert_not_called()
 
 
-def test_speech(zmq_context, mocker):
+def test_speech(mocker):
     """
     Tests the core speech actuation functionality by mocking the QI TextToSpeech
-    service and verifying that it is called correctly.
+    service and verifying that the received message is put into the queue.
     """
     mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
 
@@ -88,13 +143,406 @@ def test_speech(zmq_context, mocker):
     mock_state.qi_session = mock.Mock()
     mock_state.qi_session.service.return_value = mock_tts_service
 
-    receiver = ActuationReceiver(zmq_context)
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+
     receiver._tts_service = None
     receiver._handle_speech({"endpoint": "actuate/speech", "data": "Some message to speak."})
 
-    mock_state.qi_session.service.assert_called_once_with("ALTextToSpeech")
+    assert receiver._message_queue.qsize() == 1
 
-    mock_qi.async.assert_called_once()
+    queued_item = receiver._message_queue.get()
-    call_args = mock_qi.async.call_args[0]
+    assert queued_item == "Some message to speak."
-    assert call_args[0] == mock_tts_service.say
+
-    assert call_args[1] == "Some message to speak."
+def test_speech_priority(mocker):
+    """
+    Tests that a priority speech message is handled correctly by clearing the queue
+    and placing the priority message at the front.
+    """
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi
+
+    mock_tts_service = mock.Mock()
+    mock_state.qi_session = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_tts_service
+
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+
+    receiver._message_queue.put("old_message_1")
+    receiver._message_queue.put("old_message_2")
+
+    assert receiver._message_queue.qsize() == 2
+
+    priority_msg = {
+        "endpoint": "actuate/speech", 
+        "data": "Urgent Message",
+        "is_priority": True,
+    }
+    receiver._handle_speech(priority_msg)
+
+    assert receiver._message_queue.qsize() == 1
+    queued_item = receiver._message_queue.get()
+    assert queued_item == "Urgent Message"
+
+def test_handle_messages_loop(mocker):
+    """
+    Tests the background consumer loop (_handle_messages) processing an item.
+    Runs SYNCHRONOUSLY to ensure coverage tools pick up the lines.
+    """
+    # Patch Thread so the real background thread NEVER starts automatically
+    mocker.patch("threading.Thread")
+
+    # Mock state
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+        
+    # Setup initial speaking state to False (covers "Started speaking" print)
+    mock_state.is_speaking = False
+    
+    # Mock the TextToSpeech service
+    mock_tts_service = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_tts_service
+    
+    # Initialize receiver (Thread is patched, so no thread starts)
+    # Use Mock Context to avoid ZMQ errors
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # Manually inject service (since lazy loading might handle it, but this is safer)
+    receiver._tts_service = mock_tts_service
+
+    # This ensures the while loop iterates exactly once
+    mock_state.exit_event.is_set.side_effect = [False, True, True , True, True]
+
+    # Put an item in the queue
+    receiver._message_queue.put("Hello World")
+
+    # RUN MANUALLY in the main thread
+    # This executes the code: while -> try -> get -> if print -> speaking=True -> say
+    receiver._handle_messages()
+
+    # Assertions
+    assert receiver._message_queue.empty()
+    mock_tts_service.say.assert_called_with("Hello World")
+    assert mock_state.is_speaking is True
+
+
+def test_handle_gestures_runtime_error(mocker):
+    mocker.patch("threading.Thread")
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    
+    # Use a side_effect that returns False then True thereafter
+    mock_state.exit_event.is_set.side_effect = [False, True, True, True, True]
+    
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # ... rest of your setup ...
+    mock_anim = mock.Mock()
+    mock_anim.run.side_effect = RuntimeError("Wifi Lost")
+    receiver._animation_service = mock_anim
+    
+    receiver._gesture_queue.put("wave")
+    
+    receiver._handle_gestures()
+
+def test_handle_messages_runtime_error(mocker):
+    """
+    Tests the RuntimeError exception handler (e.g. lost WiFi connection).
+    Uses a Mock ZMQ context to avoid 'Address already in use' errors.
+    """
+    # Patch Thread so we don't accidentally spawn real threads
+    mocker.patch("threading.Thread")
+    
+    # Mock the state and logging
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+        
+    # Use a MOCK ZMQ context. 
+    # This prevents the receiver from trying to bind to a real TCP port.
+    mock_zmq_ctx = mock.Mock()
+    
+    # Initialize receiver with the mock context
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    mock_state.exit_event.is_set.side_effect = [False, True, True, True ]
+
+    receiver._message_queue.put("Test Message")
+    
+    # Setup: ...BUT the service raises RuntimeError when asked to speak
+    mock_tts = mock.Mock()
+    mock_tts.say.side_effect = RuntimeError("Connection lost")
+    receiver._tts_service = mock_tts
+    
+    # Run the loop logic manually
+    receiver._handle_messages()
+    
+    # Assertions
+    assert mock_state.exit_event.is_set.called
+
+def test_clear_queue(mocker):
+    """
+    Tests that the clear_queue method properly drains all items from the message queue.
+    """
+    mocker.patch("threading.Thread")
+    
+    # Use Mock Context
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # Populate the queue with multiple items
+    receiver._message_queue.put("msg1")
+    receiver._message_queue.put("msg2")
+    receiver._message_queue.put("msg3")
+    
+    assert receiver._message_queue.qsize() == 3
+
+    # Clear the queue
+    receiver.clear_queue()
+
+    # Assert the queue is empty
+    assert receiver._message_queue.qsize() == 0
+
+def test_gesture_no_data(mocker):
+    mock_zmq = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/single", "data": ""}, True)
+    # Just ensuring no crash
+
+
+def test_gesture_invalid_data(mocker):
+    mock_zmq = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/single", "data": 123}, True)
+    # No crash expected
+
+
+def test_gesture_single_not_found(mocker):
+    mock_zmq = mock.Mock()
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.single_gestures = ["wave", "bow"]  # allowed single gestures
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/single", "data": "unknown_gesture"}, True)
+    # No crash expected
+
+
+def test_gesture_tag_not_found(mocker):
+    mock_zmq = mock.Mock()
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.tags = ["happy", "sad"]
+
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/tag", "data": "not_a_tag"}, False)
+    # No crash expected
+
+
+def test_gesture_no_qi_session( mocker):
+    mock_zmq = mock.Mock()
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    mock_state.qi_session = None
+
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.single_gestures = ["hello"]
+
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/single", "data": "hello"}, True)
+    # No crash, path returns early
+
+
+def test_gesture_single_success(mocker):
+    mock_zmq = mock.Mock()
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    # Allow loops to run
+    mock_state.exit_event.is_set.return_value = False 
+
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi
+    
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.single_gestures = ["wave"]
+
+    mock_animation_service = mock.Mock()
+    mock_state.qi_session = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_animation_service
+
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/single", "data": "wave"}, True)
+
+    time.sleep(0.2) 
+
+    mock_state.qi_session.service.assert_called_with("ALAnimationPlayer")
+    mock_animation_service.run.assert_called_with("wave")
+
+    # CLEANUP: Signal exit AND join threads
+    mock_state.exit_event.is_set.return_value = True
+    receiver.message_thread.join(timeout=1.0)
+    receiver.gesture_thread.join(timeout=1.0)
+
+
+def test_gesture_tag_success(mocker):
+    mock_zmq = mock.Mock()
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    mock_state.exit_event.is_set.return_value = False
+
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi
+
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.tags = ["greeting"]
+
+    mock_animation_service = mock.Mock()
+    mock_state.qi_session = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_animation_service
+
+    receiver = ActuationReceiver(mock_zmq)
+    receiver._handle_gesture({"endpoint": "actuate/gesture/tag", "data": "greeting"}, False)
+
+    time.sleep(0.2)
+
+    mock_state.qi_session.service.assert_called_with("ALAnimationPlayer")
+    mock_animation_service.run.assert_called_with("greeting")
+    
+    # CLEANUP: Signal exit AND join threads
+    mock_state.exit_event.is_set.return_value = True
+    receiver.message_thread.join(timeout=1.0)
+    receiver.gesture_thread.join(timeout=1.0)
+
+
+def test_handle_message_all_routes(mocker):
+    """
+    Ensures all handle_message endpoint branches route correctly.
+    """
+    mock_zmq = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq)
+    mock_speech = mocker.patch.object(receiver, "_handle_speech")
+    mock_gesture = mocker.patch.object(receiver, "_handle_gesture")
+
+    receiver.handle_message({"endpoint": "actuate/speech", "data": "hi"})
+    receiver.handle_message({"endpoint": "actuate/gesture/tag", "data": "greeting"})
+    receiver.handle_message({"endpoint": "actuate/gesture/single", "data": "wave"})
+
+    mock_speech.assert_called_once()
+    assert mock_gesture.call_count == 2
+
+
+def test_endpoint_description(mocker):
+    mock_zmq = mock.Mock()
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.tags = ["happy"]
+    mock_tags.single_gestures = ["wave"]
+
+    receiver = ActuationReceiver(mock_zmq)
+    desc = receiver.endpoint_description()
+
+    assert "gestures" in desc
+    assert desc["gestures"] == ["happy"]
+
+    assert "single_gestures" in desc
+    assert desc["single_gestures"] == ["wave"]
+
+
+def test_gesture_single_real_gesturetags(mocker):
+    mock_zmq = mock.Mock()
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    mock_state.qi_session = mock.Mock()
+    mock_state.exit_event.is_set.return_value = False
+
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi
+
+    mock_animation_service = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_animation_service
+
+    receiver = ActuationReceiver(mock_zmq)
+
+    assert len(GestureTags.single_gestures) > 0, "GestureTags.single_gestures must not be empty"
+    gesture = GestureTags.single_gestures[0]
+
+    receiver._handle_gesture(
+        {"endpoint": "actuate/gesture/single", "data": gesture},
+        is_single=True,
+    )
+
+    time.sleep(0.2)
+    mock_state.qi_session.service.assert_called_with("ALAnimationPlayer")
+    mock_animation_service.run.assert_called_with(gesture)
+
+    # CLEANUP: Signal exit AND join threads
+    mock_state.exit_event.is_set.return_value = True
+    receiver.message_thread.join(timeout=1.0)
+    receiver.gesture_thread.join(timeout=1.0)
+
+def test_clear_gesture_queue(mocker):
+    # Prevent background threads from eating the items
+    mocker.patch("threading.Thread")
+    
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # Populate the queue
+    receiver._gesture_queue.put("gesture1")
+    receiver._gesture_queue.put("gesture2")
+    
+    assert receiver._gesture_queue.qsize() == 2
+
+    # Clear the queue
+    receiver.clear_gesture_queue()
+
+    # Assert the queue is empty
+    assert receiver._gesture_queue.qsize() == 0
+
+
+def test_gesture_priority_clears_queue(mocker):
+    mocker.patch("threading.Thread")
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    
+    # Mock QI and Tags so valid checks pass
+    mock_qi = mock.Mock()
+    sys.modules["qi"] = mock_qi
+    mock_tags = mocker.patch("robot_interface.endpoints.actuation_receiver.GestureTags")
+    mock_tags.single_gestures = ["urgent_wave"]
+    
+    # Setup Animation Service
+    mock_anim = mock.Mock()
+    mock_state.qi_session.service.return_value = mock_anim
+
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # Pre-fill queue with "slow" gestures
+    receiver._gesture_queue.put("slow_gesture_1")
+    receiver._gesture_queue.put("slow_gesture_2")
+    
+    assert receiver._gesture_queue.qsize() == 2
+
+    # Send priority gesture
+    priority_msg = {
+        "endpoint": "actuate/gesture/single", 
+        "data": "urgent_wave",
+        "is_priority": True,
+    }
+    receiver._handle_gesture(priority_msg, is_single=True)
+
+    # Assert old items are gone and only new one remains
+    assert receiver._gesture_queue.qsize() == 1
+    assert receiver._gesture_queue.get() == "urgent_wave"
+
+
+def test_handle_gestures_loop_empty(mocker):
+    mocker.patch("threading.Thread")
+    mock_state = mocker.patch("robot_interface.endpoints.actuation_receiver.state")
+    
+    mock_zmq_ctx = mock.Mock()
+    receiver = ActuationReceiver(mock_zmq_ctx)
+    
+    # Run loop exactly once
+    mock_state.exit_event.is_set.side_effect = [False, True, True, True]
+    
+    # We don't put anything in the queue, so .get(timeout=0.1) will raise Queue.Empty.
+    # The code should catch it and pass.
+    receiver._handle_gestures()
+    
+    # If we reached here without raising an exception, the test passes.
+    # We can assert that the queue is still valid/empty.
+    assert receiver._gesture_queue.empty()

test/unit/test_get_config.py

@@ -0,0 +1,45 @@
+from robot_interface.utils.get_config import get_config
+
+
+def test_get_config_prefers_explicit_value(monkeypatch):
+    """
+    When a direct value is provided it should be returned without reading the environment.
+    """
+    monkeypatch.setenv("GET_CONFIG_TEST", "from-env")
+
+    result = get_config("explicit", "GET_CONFIG_TEST", "default")
+
+    assert result == "explicit"
+
+
+def test_get_config_returns_env_value(monkeypatch):
+    """
+    If value is None the environment variable should be used.
+    """
+    monkeypatch.setenv("GET_CONFIG_TEST", "from-env")
+
+    result = get_config(None, "GET_CONFIG_TEST", "default")
+
+    assert result == "from-env"
+
+
+def test_get_config_casts_env_value(monkeypatch):
+    """
+    The env value should be cast when a cast function is provided.
+    """
+    monkeypatch.setenv("GET_CONFIG_PORT", "1234")
+
+    result = get_config(None, "GET_CONFIG_PORT", 0, int)
+
+    assert result == 1234
+
+
+def test_get_config_casts_default_when_env_missing(monkeypatch):
+    """
+    When the env var is missing it should fall back to the default and still apply the cast.
+    """
+    monkeypatch.delenv("GET_CONFIG_MISSING", raising=False)
+
+    result = get_config(None, "GET_CONFIG_MISSING", "42", int)
+
+    assert result == 42