diff --git a/.github/workflows/scripts-test.yaml b/.github/workflows/scripts-test.yaml
index e11424924..9c7f1ace5 100644
--- a/.github/workflows/scripts-test.yaml
+++ b/.github/workflows/scripts-test.yaml
@@ -48,6 +48,9 @@ jobs:
       
       - name: Check code health
         run: python $scripts_validation_dir/code_health.py -i .
+      
+      - name: Check code documentation
+        run: python $scripts_validation_dir/doc_style.py -i .
 
   test:
     name: Test scripts
diff --git a/benchmark/flake_rules.json b/benchmark/validation_rules.json
similarity index 74%
rename from benchmark/flake_rules.json
rename to benchmark/validation_rules.json
index 3a98aec8a..97be3cf88 100644
--- a/benchmark/flake_rules.json
+++ b/benchmark/validation_rules.json
@@ -9,9 +9,18 @@
       "src/components/pytorch_image_classifier/torch_helper/model/__init__.py:F401",
       "src/components/tensorflow_image_segmentation/train.py:E402"
     ],
+    "max-line-length": 200,
     "justifications": [
       "__init__.py - namespace level files excluded from validation",
       "E402 - allow module level imports for specific files"
     ]
+  },
+  "doc": {
+    "exclude": [
+      "."
+    ],
+    "justifications": [
+      ". - not enabled yet"
+    ]
   }
-}
\ No newline at end of file
+}
diff --git a/responsibleai/validation_rules.json b/responsibleai/validation_rules.json
new file mode 100644
index 000000000..174c66ceb
--- /dev/null
+++ b/responsibleai/validation_rules.json
@@ -0,0 +1,13 @@
+{
+  "pep8": {
+    "max-line-length": 200
+  },
+  "doc": {
+    "exclude": [
+      "."
+    ],
+    "justifications": [
+      ". - not enabled yet"
+    ]
+  }
+}
diff --git a/scripts/azureml-assets/azureml/assets/config.py b/scripts/azureml-assets/azureml/assets/config.py
index b16e6f408..ca781b343 100644
--- a/scripts/azureml-assets/azureml/assets/config.py
+++ b/scripts/azureml-assets/azureml/assets/config.py
@@ -173,7 +173,7 @@ class EnvironmentConfig(Config):
     Example:
 
     image:
-      name: azureml/curated/tensorflow-2.7-ubuntu20.04-py38-cuda11-gpu # Can include registry hostname and template tags
+      name: azureml/curated/tensorflow-2.7-ubuntu20.04-py38-cuda11-gpu # Can include registry hostname & template tags
       os: linux
       context: # If not specified, image won't be built
         dir: context
@@ -414,7 +414,8 @@ class AssetConfig(Config):
         if not Config._is_set(name):
             name = self.spec_as_object().name
             if Config._contains_template(name):
-                raise ValidationException(f"Tried to read asset name from spec, but it includes a template tag: {name}")
+                raise ValidationException(f"Tried to read asset name from spec, "
+                                          f"but it includes a template tag: {name}")
         return name
 
     @property
@@ -439,7 +440,8 @@ class AssetConfig(Config):
                 if self.auto_version:
                     version = None
                 else:
-                    raise ValidationException(f"Tried to read asset version from spec, but it includes a template tag: {version}")
+                    raise ValidationException(f"Tried to read asset version from spec, "
+                                              f"but it includes a template tag: {version}")
         return str(version) if version is not None else None
 
     @property
diff --git a/scripts/azureml-assets/azureml/assets/copy_unreleased_assets.py b/scripts/azureml-assets/azureml/assets/copy_unreleased_assets.py
index 90713d669..10113fc7e 100644
--- a/scripts/azureml-assets/azureml/assets/copy_unreleased_assets.py
+++ b/scripts/azureml-assets/azureml/assets/copy_unreleased_assets.py
@@ -49,9 +49,12 @@ def copy_unreleased_assets(release_directory_root: Path,
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-r", "--release-directory", required=True, type=Path, help="Directory to which the release branch has been cloned")
-    parser.add_argument("-o", "--output-directory", required=True, type=Path, help="Directory to which unreleased assets will be written")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
+    parser.add_argument("-r", "--release-directory", required=True, type=Path,
+                        help="Directory to which the release branch has been cloned")
+    parser.add_argument("-o", "--output-directory", required=True, type=Path,
+                        help="Directory to which unreleased assets will be written")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
     args = parser.parse_args()
 
     # Release assets
diff --git a/scripts/azureml-assets/azureml/assets/environment/build.py b/scripts/azureml-assets/azureml/assets/environment/build.py
index 7d72ace40..eee147847 100644
--- a/scripts/azureml-assets/azureml/assets/environment/build.py
+++ b/scripts/azureml-assets/azureml/assets/environment/build.py
@@ -139,12 +139,14 @@ def build_images(input_dirs: List[Path],
     with ThreadPoolExecutor(max_parallel) as pool:
         # Find environments under image root directories
         futures = []
-        for asset_config in util.find_assets(input_dirs, asset_config_filename, assets.AssetType.ENVIRONMENT, changed_files):
+        for asset_config in util.find_assets(input_dirs, asset_config_filename, assets.AssetType.ENVIRONMENT,
+                                             changed_files):
             env_config = asset_config.environment_config_as_object()
 
             # Filter by OS
             if os_to_build and env_config.os.value != os_to_build:
-                logger.print(f"Skipping build of image for {asset_config.name}: Operating system {env_config.os.value} != {os_to_build}")
+                logger.print(f"Skipping build of image for {asset_config.name}: "
+                             f"Operating system {env_config.os.value} != {os_to_build}")
                 continue
 
             # Pin versions
@@ -162,7 +164,8 @@ def build_images(input_dirs: List[Path],
 
                 # Copy file to output directory without building
                 if output_directory:
-                    util.copy_asset_to_output_dir(asset_config=asset_config, output_directory=output_directory, add_subdir=True)
+                    util.copy_asset_to_output_dir(asset_config=asset_config, output_directory=output_directory,
+                                                  add_subdir=True)
                 continue
 
             # Tag with version from spec
@@ -185,13 +188,15 @@ def build_images(input_dirs: List[Path],
             logger.print(output)
             logger.end_group()
             if return_code != 0:
-                logger.log_error(f"Build of image for {asset_config.name} failed with exit status {return_code}", "Build failure")
+                logger.log_error(f"Build of image for {asset_config.name} failed with exit status {return_code}",
+                                 "Build failure")
                 counters[FAILED_COUNT] += 1
             else:
                 logger.log_debug(f"Successfully built image for {asset_config.name}")
                 counters[SUCCESS_COUNT] += 1
                 if output_directory:
-                    util.copy_asset_to_output_dir(asset_config=asset_config, output_directory=output_directory, add_subdir=True)
+                    util.copy_asset_to_output_dir(asset_config=asset_config, output_directory=output_directory,
+                                                  add_subdir=True)
 
     # Set variables
     for counter_name in COUNTERS:
@@ -206,19 +211,32 @@ def build_images(input_dirs: List[Path],
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dirs", required=True, help="Comma-separated list of directories containing environments to build")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
-    parser.add_argument("-o", "--output-directory", type=Path, help="Directory to which successfully built environments will be written")
-    parser.add_argument("-l", "--build-logs-dir", required=True, type=Path, help="Directory to receive build logs")
-    parser.add_argument("-p", "--max-parallel", type=int, default=25, help="Maximum number of images to build at the same time")
-    parser.add_argument("-c", "--changed-files", help="Comma-separated list of changed files, used to filter images")
-    parser.add_argument("-O", "--os-to-build", choices=[i.value for i in list(assets.Os)], help="Only build environments based on this OS")
-    parser.add_argument("-r", "--registry", help="Container registry on which to build images")
-    parser.add_argument("-g", "--resource-group", help="Resource group containing the container registry")
-    parser.add_argument("-P", "--pin-versions", action="store_true", help="Pin images/packages to latest versions")
-    parser.add_argument("-t", "--tag-with-version", action="store_true", help="Tag image names using the version in the asset's spec file")
-    parser.add_argument("-T", "--test-command", help="If building on ACR, command used to test image, relative to build context root")
-    parser.add_argument("-u", "--push", action="store_true", help="If building on ACR, push after building and (optionally) testing")
+    parser.add_argument("-i", "--input-dirs", required=True,
+                        help="Comma-separated list of directories containing environments to build")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
+    parser.add_argument("-o", "--output-directory", type=Path,
+                        help="Directory to which successfully built environments will be written")
+    parser.add_argument("-l", "--build-logs-dir", required=True, type=Path,
+                        help="Directory to receive build logs")
+    parser.add_argument("-p", "--max-parallel", type=int, default=25,
+                        help="Maximum number of images to build at the same time")
+    parser.add_argument("-c", "--changed-files",
+                        help="Comma-separated list of changed files, used to filter images")
+    parser.add_argument("-O", "--os-to-build", choices=[i.value for i in list(assets.Os)],
+                        help="Only build environments based on this OS")
+    parser.add_argument("-r", "--registry",
+                        help="Container registry on which to build images")
+    parser.add_argument("-g", "--resource-group",
+                        help="Resource group containing the container registry")
+    parser.add_argument("-P", "--pin-versions", action="store_true",
+                        help="Pin images/packages to latest versions")
+    parser.add_argument("-t", "--tag-with-version", action="store_true",
+                        help="Tag image names using the version in the asset's spec file")
+    parser.add_argument("-T", "--test-command",
+                        help="If building on ACR, command used to test image, relative to build context root")
+    parser.add_argument("-u", "--push", action="store_true",
+                        help="If building on ACR, push after building and (optionally) testing")
     args = parser.parse_args()
 
     # Ensure dependent args are present
diff --git a/scripts/azureml-assets/azureml/assets/environment/create_environments.py b/scripts/azureml-assets/azureml/assets/environment/create_environments.py
deleted file mode 100644
index ce4e8f682..000000000
--- a/scripts/azureml-assets/azureml/assets/environment/create_environments.py
+++ /dev/null
@@ -1,99 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-
-import argparse
-import json
-import requests
-from pathlib import Path
-from urllib.parse import urljoin
-
-import azureml.assets.util as util
-from azureml.assets.util import logger
-
-ENV_URL_TEMPLATE = "environment/1.0/environments/{environment}/versions/{version}/registry/{registry}"
-
-
-def create_environment(base_url: str,
-                       name: str,
-                       version: str,
-                       registry: str,
-                       image_name: str,
-                       access_token: str,
-                       env_def_with_metadata: object):
-    url = urljoin(base_url, ENV_URL_TEMPLATE.format(environment=name, version=version, registry=registry))
-    headers = {
-        'Authorization': f"Bearer {access_token}",
-        'Content-Type': "application/json"
-    }
-    params = {'imageName': image_name}
-    response = requests.put(url, headers=headers, json=env_def_with_metadata, params=params)
-    if not response.ok:
-        logger.log_error(f"Failed to create {name} version {version}: {response.status_code} {response.text}")
-    return response.ok
-
-
-def create_environments(deployment_config_file_path: Path,
-                        base_url: str,
-                        registry: str,
-                        access_token: str,
-                        version_template: str = None,
-                        tag_template: str = None):
-    # Load config
-    base_path = deployment_config_file_path.parent
-    with open(deployment_config_file_path) as f:
-        deployment_config = json.loads(f.read())
-
-    # Iterate over environments
-    errors = False
-    for name, values in deployment_config.items():
-        logger.print(f"Creating environment {name} in {registry}")
-
-        # Coerce values into a list
-        values_list = values if isinstance(values, list) else [values]
-
-        # Iterate over versions, although there's likely just one
-        for value in values_list:
-            version = value['version']
-
-            # Load EnvironmentDefinitionWithSetMetadataDto
-            with open(base_path / value['path']) as f:
-                env_def_with_metadata = json.loads(f.read())
-
-            # Apply tag template to image name
-            full_image_name = util.apply_tag_template(value['publish']['fullImageName'], tag_template)
-
-            # Apply version template
-            version = util.apply_version_template(version, version_template)
-
-            # Create environment
-            success = create_environment(base_url=base_url, name=name, version=version, registry=registry,
-                                         image_name=full_image_name, access_token=access_token,
-                                         env_def_with_metadata=env_def_with_metadata)
-            if not success:
-                errors = True
-
-    # Final messages
-    if errors:
-        raise Exception("Errors occurred while creating environments")
-
-
-if __name__ == "__main__":
-    # Handle command-line args
-    parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--deployment-config", required=True, type=Path, help="Path to deployment config file")
-    parser.add_argument("-u", "--url", required=True, help="Base AzureML URL, example: https://master.api.azureml-test.ms")
-    parser.add_argument("-r", "--registry", required=True, help="Name of the target registry")
-    parser.add_argument("-t", "--token", required=True, help="Access token to use for bearer authentication")
-    parser.add_argument("-v", "--version-template", help="Template to apply to the version from the deployment config "
-                        "file, example: '{version}.dev1'")
-    parser.add_argument("-T", "--tag-template", help="Template to apply to fullImageName tags from the deployment "
-                        "config file, example: '{tag}.dev1'")
-    args = parser.parse_args()
-
-    create_environments(deployment_config_file_path=args.deployment_config,
-                        base_url=args.url,
-                        registry=args.registry,
-                        access_token=args.token,
-                        version_template=args.version_template,
-                        tag_template=args.tag_template)
diff --git a/scripts/azureml-assets/azureml/assets/environment/pin_image_versions.py b/scripts/azureml-assets/azureml/assets/environment/pin_image_versions.py
index 0a169f082..e0300e103 100644
--- a/scripts/azureml-assets/azureml/assets/environment/pin_image_versions.py
+++ b/scripts/azureml-assets/azureml/assets/environment/pin_image_versions.py
@@ -126,8 +126,10 @@ def transform_file(input_file: Path, output_file: Path = None):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input", type=Path, help="File containing images to pin to latest versions", required=True)
-    parser.add_argument("-o", "--output", type=Path, help="File to which output will be written. Defaults to the input file if not specified.")
+    parser.add_argument("-i", "--input", type=Path,
+                        help="File containing images to pin to latest versions", required=True)
+    parser.add_argument("-o", "--output", type=Path,
+                        help="File to which output will be written. Defaults to the input file.")
     args = parser.parse_args()
 
     output = args.output or args.input
diff --git a/scripts/azureml-assets/azureml/assets/environment/pin_package_versions.py b/scripts/azureml-assets/azureml/assets/environment/pin_package_versions.py
index 481f72c28..3d49fd48b 100644
--- a/scripts/azureml-assets/azureml/assets/environment/pin_package_versions.py
+++ b/scripts/azureml-assets/azureml/assets/environment/pin_package_versions.py
@@ -103,8 +103,10 @@ def transform_file(input_file: Path, output_file: Path = None):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input", type=Path, help="File containing packages to pin to latest versions", required=True)
-    parser.add_argument("-o", "--output", type=Path, help="File to which output will be written. Defaults to the input file if not specified.")
+    parser.add_argument("-i", "--input", type=Path,
+                        help="File containing packages to pin to latest versions", required=True)
+    parser.add_argument("-o", "--output", type=Path,
+                        help="File to which output will be written. Defaults to the input file.")
     args = parser.parse_args()
 
     output = args.output or args.input
diff --git a/scripts/azureml-assets/azureml/assets/environment/pin_versions.py b/scripts/azureml-assets/azureml/assets/environment/pin_versions.py
index 5fcd1ba4f..51cda9aea 100644
--- a/scripts/azureml-assets/azureml/assets/environment/pin_versions.py
+++ b/scripts/azureml-assets/azureml/assets/environment/pin_versions.py
@@ -30,8 +30,10 @@ def transform_file(input_file: Path, output_file: Path = None):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input", type=Path, help="File containing images/packages to pin to latest versions", required=True)
-    parser.add_argument("-o", "--output", type=Path, help="File to which output will be written. Defaults to the input file if not specified.")
+    parser.add_argument("-i", "--input", type=Path,
+                        help="File containing images/packages to pin to latest versions", required=True)
+    parser.add_argument("-o", "--output", type=Path,
+                        help="File to which output will be written. Defaults to the input file.")
     args = parser.parse_args()
 
     output = args.output or args.input
diff --git a/scripts/azureml-assets/azureml/assets/environment/push.py b/scripts/azureml-assets/azureml/assets/environment/push.py
deleted file mode 100644
index d521a0ffe..000000000
--- a/scripts/azureml-assets/azureml/assets/environment/push.py
+++ /dev/null
@@ -1,86 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-
-import argparse
-import sys
-from concurrent.futures import as_completed, ThreadPoolExecutor
-from datetime import datetime, timedelta, timezone
-from subprocess import run, PIPE, STDOUT
-from timeit import default_timer as timer
-from typing import List
-
-from azureml.assets.util import logger
-
-TAG_DATETIME_FORMAT = "%Y%m%d%H%M%S"
-
-
-def tag_image(image_name: str, target_image_name: str):
-    p = run(["docker", "tag", image_name, target_image_name],
-            stdout=PIPE,
-            stderr=STDOUT)
-    return (p.returncode, p.stdout.decode())
-
-
-def push_image(image_name: str, all_tags: bool = False):
-    # Create args
-    args = ["docker", "push"]
-    if all_tags:
-        args.append("--all-tags")
-    args.append(image_name)
-
-    # Push
-    logger.print(f"Pushing {image_name}")
-    start = timer()
-    p = run(args,
-            stdout=PIPE,
-            stderr=STDOUT)
-    end = timer()
-    logger.print(f"{image_name} pushed in {timedelta(seconds=end-start)}")
-    return (image_name, p.returncode, p.stdout.decode())
-
-
-def push_images(image_names: List[str], target_image_prefix: str, tags: List[str], max_parallel: int):
-    with ThreadPoolExecutor(max_parallel) as pool:
-        futures = []
-        for image_name in image_names:
-            # Tag image
-            target_image_base_name = f"{target_image_prefix}{image_name}"
-            logger.print(f"Tagging {target_image_base_name} as {tags}")
-            for tag in tags:
-                target_image_name = f"{target_image_base_name}:{tag}"
-                (return_code, output) = tag_image(image_name, target_image_name)
-                if return_code != 0:
-                    logger.log_error(f"Failed to tag {image_name} as {target_image_name}: {output}", "Tag failure")
-                    sys.exit(1)
-
-            # Start push
-            futures.append(pool.submit(push_image, target_image_base_name, True))
-
-        for future in as_completed(futures):
-            (image_name, return_code, output) = future.result()
-            logger.start_group(f"{image_name} push log")
-            logger.print(output)
-            logger.end_group()
-            if return_code != 0:
-                logger.log_error(f"Push of {image_name} failed with exit status {return_code}", "Push failure")
-                sys.exit(1)
-
-
-if __name__ == '__main__':
-    # Handle command-line args
-    parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--image-names", required=True, help="Comma-separated list of image names to push")
-    parser.add_argument("-t", "--target-image-prefix", required=True, help="Prefix to use when tagging images")
-    parser.add_argument("-p", "--max-parallel", type=int, default=5, help="Maximum number of images to push at the same time")
-    args = parser.parse_args()
-
-    # Convert comma-separated values to lists
-    image_names = args.image_names.split(",") if args.image_names else []
-
-    # Generate tags
-    timestamp_tag = datetime.now(timezone.utc).strftime(TAG_DATETIME_FORMAT)
-    tags = [timestamp_tag, "latest"]
-
-    # Push images
-    push_images(image_names, args.target_image_prefix, tags, args.max_parallel)
diff --git a/scripts/azureml-assets/azureml/assets/environment/test_images.py b/scripts/azureml-assets/azureml/assets/environment/test_images.py
deleted file mode 100644
index c891ed4e3..000000000
--- a/scripts/azureml-assets/azureml/assets/environment/test_images.py
+++ /dev/null
@@ -1,92 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-
-import argparse
-import sys
-from collections import Counter
-from datetime import timedelta
-from pathlib import Path
-from subprocess import run, PIPE, STDOUT
-from timeit import default_timer as timer
-from typing import List, Tuple
-
-import azureml.assets as assets
-import azureml.assets.util as util
-from azureml.assets.util import logger
-
-SUCCESS_COUNT = "success_count"
-FAILED_COUNT = "failed_count"
-COUNTERS = [SUCCESS_COUNT, FAILED_COUNT]
-TEST_PHRASE = "hello world!"
-
-
-def test_image(asset_config: assets.AssetConfig, image_name: str) -> Tuple[int, str]:
-    logger.print(f"Testing image for {asset_config.name}")
-    start = timer()
-    p = run(["docker", "run", "--entrypoint", "python", image_name, "-c", f"print(\"{TEST_PHRASE}\")"],
-            stdout=PIPE,
-            stderr=STDOUT)
-    end = timer()
-    logger.print(f"Image for {asset_config.name} tested in {timedelta(seconds=end-start)}")
-    return (p.returncode, p.stdout.decode())
-
-
-def test_images(input_dirs: List[Path],
-                asset_config_filename: str,
-                output_directory: Path,
-                os_to_test: str = None) -> bool:
-    counters = Counter()
-    for asset_config in util.find_assets(input_dirs, asset_config_filename, assets.AssetType.ENVIRONMENT):
-        env_config = asset_config.environment_config_as_object()
-
-        # Filter by OS
-        if os_to_test and env_config.os.value != os_to_test:
-            logger.print(f"Not testing image for {asset_config.name}: Operating system {env_config.os.value} != {os_to_test}")
-            continue
-
-        # Skip images without build context
-        if not env_config.build_enabled:
-            logger.print(f"Not testing image for {asset_config.name}: No build context specified")
-            continue
-
-        # Test image
-        return_code, output = test_image(asset_config, env_config.image_name)
-        if return_code != 0 or not output.startswith(TEST_PHRASE):
-            logger.log_error(f"Testing of image for {asset_config.name} failed: {output}", title="Testing failure")
-            counters[FAILED_COUNT] += 1
-        else:
-            logger.log_debug(f"Successfully tested image for {asset_config.name}")
-            counters[SUCCESS_COUNT] += 1
-            if output_directory:
-                util.copy_asset_to_output_dir(asset_config=asset_config, output_dir=output_directory, add_subdir=True)
-
-    # Set variables
-    for counter_name in COUNTERS:
-        logger.set_output(counter_name, counters[counter_name])
-
-    if counters[FAILED_COUNT] > 0:
-        logger.log_error(f"{counters[FAILED_COUNT]} environment image(s) failed to test")
-        return False
-    return True
-
-
-if __name__ == '__main__':
-    # Handle command-line args
-    parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dirs", required=True, help="Comma-separated list of directories containing environments to test")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
-    parser.add_argument("-o", "--output-directory", type=Path, help="Directory to which successfully tested environments will be written")
-    parser.add_argument("-O", "--os-to-test", choices=[i.value for i in list(assets.Os)], help="Only test environments based on this OS")
-    args = parser.parse_args()
-
-    # Convert comma-separated values to lists
-    input_dirs = [Path(d) for d in args.input_dirs.split(",")]
-
-    # Test images
-    success = test_images(input_dirs=input_dirs,
-                          asset_config_filename=args.asset_config_filename,
-                          output_directory=args.output_directory,
-                          os_to_test=args.os_to_test)
-    if not success:
-        sys.exit(1)
diff --git a/scripts/azureml-assets/azureml/assets/release/asset_publish.py b/scripts/azureml-assets/azureml/assets/release/asset_publish.py
index 1bf0ca7e0..21f907a51 100644
--- a/scripts/azureml-assets/azureml/assets/release/asset_publish.py
+++ b/scripts/azureml-assets/azureml/assets/release/asset_publish.py
@@ -81,7 +81,8 @@ if __name__ == '__main__':
         if registry_name != "azureml":
             final_version = final_version + '-' + component_version_with_buildId
         print("final version: "+final_version)
-        cmd = f"az ml component create --file {spec_path} --registry-name {registry_name} --version {final_version} --workspace {workspace} --resource-group {resource_group}"
+        cmd = f"az ml component create --file {spec_path} --registry-name {registry_name} --version {final_version} " \
+              f"--workspace {workspace} --resource-group {resource_group}"
         print(cmd)
         try:
             check_call(cmd, shell=True)
diff --git a/scripts/azureml-assets/azureml/assets/release/e2e_test.py b/scripts/azureml-assets/azureml/assets/release/e2e_test.py
index dd091a8c1..483064b45 100644
--- a/scripts/azureml-assets/azureml/assets/release/e2e_test.py
+++ b/scripts/azureml-assets/azureml/assets/release/e2e_test.py
@@ -52,7 +52,8 @@ if __name__ == '__main__':
             data = yaml.load(fp, Loader=yaml.FullLoader)
             for test_group in data:
                 print(f"now processing test group: {test_group}")
-                p = run(f"python3 -u group_test.py -i {area} -g {test_group} -s {subscription_id} -r {resource_group} -w {workspace}", shell=True)
+                p = run(f"python3 -u group_test.py -i {area} -g {test_group} -s {subscription_id} -r {resource_group} "
+                        f"-w {workspace}", shell=True)
                 return_code = p.returncode
                 print(return_code)
                 final_report[area.name].append(f"test group {test_group} returned {return_code}")
diff --git a/scripts/azureml-assets/azureml/assets/release/test_file_convert.py b/scripts/azureml-assets/azureml/assets/release/test_file_convert.py
index 03caaa46e..df9fa8832 100644
--- a/scripts/azureml-assets/azureml/assets/release/test_file_convert.py
+++ b/scripts/azureml-assets/azureml/assets/release/test_file_convert.py
@@ -18,9 +18,12 @@ def copy_replace_dir(source: Path, dest: Path):
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dir", required=True, type=Path, help="dir path of tests.yml")
-    parser.add_argument("-a", "--test-area", required=True, type=str, help="the test area name")
-    parser.add_argument("-r", "--release-directory", required=True, type=Path, help="Directory to which the release branch has been cloned")
+    parser.add_argument("-i", "--input-dir", required=True, type=Path,
+                        help="dir path of tests.yml")
+    parser.add_argument("-a", "--test-area", required=True, type=str,
+                        help="the test area name")
+    parser.add_argument("-r", "--release-directory", required=True, type=Path,
+                        help="Directory to which the release branch has been cloned")
     args = parser.parse_args()
     yaml_name = "tests.yml"
     tests_folder = args.release_directory / "tests" / args.test_area
diff --git a/scripts/azureml-assets/azureml/assets/tag_released_assets.py b/scripts/azureml-assets/azureml/assets/tag_released_assets.py
index c7e2d6bdd..b4aea70f2 100644
--- a/scripts/azureml-assets/azureml/assets/tag_released_assets.py
+++ b/scripts/azureml-assets/azureml/assets/tag_released_assets.py
@@ -42,9 +42,12 @@ def tag_released_assets(input_directory: Path,
 if __name__ == "__main__":
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-directory", required=True, type=Path, help="Directory containing released assets")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
-    parser.add_argument("-r", "--release-directory", required=True, type=Path, help="Directory to which the release branch has been cloned")
+    parser.add_argument("-i", "--input-directory", required=True, type=Path,
+                        help="Directory containing released assets")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
+    parser.add_argument("-r", "--release-directory", required=True, type=Path,
+                        help="Directory to which the release branch has been cloned")
     parser.add_argument("-u", "--username", help="Username for git push")
     parser.add_argument("-e", "--email", help="Email for git push")
     args = parser.parse_args()
diff --git a/scripts/azureml-assets/azureml/assets/test_assets.py b/scripts/azureml-assets/azureml/assets/test_assets.py
index 71e2880c2..e685acebc 100644
--- a/scripts/azureml-assets/azureml/assets/test_assets.py
+++ b/scripts/azureml-assets/azureml/assets/test_assets.py
@@ -105,9 +105,12 @@ def test_assets(input_dirs: List[Path],
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dirs", required=True, help="Comma-separated list of directories containing environments to test")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
-    parser.add_argument("-p", "--package-versions-file", required=True, type=Path, help="File with package versions for the base conda environment")
+    parser.add_argument("-i", "--input-dirs", required=True,
+                        help="Comma-separated list of directories containing environments to test")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
+    parser.add_argument("-p", "--package-versions-file", required=True, type=Path,
+                        help="File with package versions for the base conda environment")
     parser.add_argument("-c", "--changed-files", help="Comma-separated list of changed files, used to filter assets")
     parser.add_argument("-r", "--reports-dir", type=Path, help="Directory for pytest reports")
     args = parser.parse_args()
diff --git a/scripts/azureml-assets/azureml/assets/update_assets.py b/scripts/azureml-assets/azureml/assets/update_assets.py
index 96f405dc0..ef4c7c96c 100644
--- a/scripts/azureml-assets/azureml/assets/update_assets.py
+++ b/scripts/azureml-assets/azureml/assets/update_assets.py
@@ -171,12 +171,18 @@ def update_assets(input_dirs: List[Path],
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dirs", required=True, help="Comma-separated list of directories containing assets")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
-    parser.add_argument("-r", "--release-directory", required=True, type=Path, help="Directory to which the release branch has been cloned")
-    parser.add_argument("-o", "--output-directory", type=Path, help="Directory to which new/updated assets will be written, defaults to release directory")
-    parser.add_argument("-c", "--copy-only", action="store_true", help="Just copy assets into the release directory")
-    parser.add_argument("-s", "--skip-unreleased", action="store_true", help="Skip unreleased dynamically-versioned assets in the release branch")
+    parser.add_argument("-i", "--input-dirs", required=True,
+                        help="Comma-separated list of directories containing assets")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
+    parser.add_argument("-r", "--release-directory", required=True, type=Path,
+                        help="Directory to which the release branch has been cloned")
+    parser.add_argument("-o", "--output-directory", type=Path,
+                        help="Directory to which new/updated assets will be written, defaults to release directory")
+    parser.add_argument("-c", "--copy-only", action="store_true",
+                        help="Just copy assets into the release directory")
+    parser.add_argument("-s", "--skip-unreleased", action="store_true",
+                        help="Skip unreleased dynamically-versioned assets in the release branch")
     args = parser.parse_args()
 
     # Convert comma-separated values to lists
diff --git a/scripts/azureml-assets/azureml/assets/update_spec.py b/scripts/azureml-assets/azureml/assets/update_spec.py
index 66af522dc..b7be89975 100644
--- a/scripts/azureml-assets/azureml/assets/update_spec.py
+++ b/scripts/azureml-assets/azureml/assets/update_spec.py
@@ -67,9 +67,11 @@ def update(asset_config: assets.AssetConfig, release_directory_root: Path = None
     Args:
         asset_config (assets.AssetConfig): AssetConfig object.
         release_directory_root (Path, optional): Directory to which the release branch has been cloned.
-        output_file (Path, optional): File to which updated spec file will be written. If unspecified, the original spec file will be updated.
+        output_file (Path, optional): File to which updated spec file will be written.
+                                      If unspecified, the original spec file will be updated.
         version (str, optional): Version to use instead of the one in the asset config file.
-        include_commit_hash (bool, optional): Whether to include the commit hash in the data available for template replacemnt.
+        include_commit_hash (bool, optional): Whether to include the commit hash in the data available for
+                                              template replacemnt.
         data (Dict[str, object], optional): If provided, use this data instead of calling create_template_data().
     """
     # Reuse or create data
@@ -92,9 +94,12 @@ def update(asset_config: assets.AssetConfig, release_directory_root: Path = None
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument("-a", "--asset-config", required=True, type=Path, help="Asset config file that points to the spec file to update")
-    parser.add_argument("-r", "--release-directory", type=Path, help="Directory to which the release branch has been cloned")
-    parser.add_argument("-o", "--output", type=Path, help="File to which output will be written. Defaults to the original spec file if not specified.")
+    parser.add_argument("-a", "--asset-config", required=True, type=Path,
+                        help="Asset config file that points to the spec file to update")
+    parser.add_argument("-r", "--release-directory", type=Path,
+                        help="Directory to which the release branch has been cloned")
+    parser.add_argument("-o", "--output", type=Path,
+                        help="File to which output will be written. Defaults to the original spec file.")
     args = parser.parse_args()
 
     update(asset_config=args.asset_config, release_directory_root=args.release_directory, output_file=args.output)
diff --git a/scripts/azureml-assets/azureml/assets/util/util.py b/scripts/azureml-assets/azureml/assets/util/util.py
index c9c07f159..37a8432b0 100644
--- a/scripts/azureml-assets/azureml/assets/util/util.py
+++ b/scripts/azureml-assets/azureml/assets/util/util.py
@@ -15,7 +15,7 @@ RELEASE_SUBDIR = "latest"
 EXCLUDE_DIR_PREFIX = "!"
 
 
-# See https://stackoverflow.com/questions/4187564/recursively-compare-two-directories-to-ensure-they-have-the-same-files-and-subdi
+# See https://stackoverflow.com/questions/4187564
 def are_dir_trees_equal(dir1: Path, dir2: Path, enable_logging: bool = False, ignore_eol: bool = True) -> bool:
     """Compare two directories recursively based on files names and content.
 
@@ -29,7 +29,6 @@ def are_dir_trees_equal(dir1: Path, dir2: Path, enable_logging: bool = False, ig
         bool: True if the directory trees are the same and there were no errors
             while accessing the directories or files, False otherwise.
     """
-
     dirs_cmp = filecmp.dircmp(dir1, dir2)
     if dirs_cmp.left_only:
         _log_diff(f"Compared {dir1} and {dir2} and found these only in {dir1}: {dirs_cmp.left_only}", enable_logging)
@@ -241,8 +240,8 @@ def find_assets(input_dirs: Union[List[Path], Path],
     Args:
         input_dirs (Union[List[Path], Path]): Directories to search in.
         asset_config_filename (str, optional): Asset config filename to search for.
-        types (Union[List[assets.AssetType], assets.AssetType], optional): AssetTypes to search for. Will not filter if unspecified.
-        changed_files (List[Path], optional): Changed files, used to filter assets in input_dirs. Will not filter if unspecified.
+        types (Union[List[assets.AssetType], assets.AssetType], optional): AssetTypes to search for.
+        changed_files (List[Path], optional): Changed files, used to filter assets in input_dirs.
         exclude_dirs (Union[List[Path], Path], optional): Directories that should be excluded from the search.
 
     Returns:
@@ -273,7 +272,7 @@ def find_asset_config_files(input_dirs: Union[List[Path], Path],
     Args:
         input_dirs (Union[List[Path], Path]): Directories to search in.
         asset_config_filename (str): Asset config filename to search for.
-        changed_files (List[Path], optional): Changed files, used to filter assets in input_dirs. Will not filter if unspecified.
+        changed_files (List[Path], optional): Changed files, used to filter assets in input_dirs.
         exclude_dirs (Union[List[Path], Path], optional): Directories that should be excluded from the search.
 
     Returns:
diff --git a/scripts/azureml-assets/azureml/assets/validate_assets.py b/scripts/azureml-assets/azureml/assets/validate_assets.py
index 092b92e26..ab145ccec 100644
--- a/scripts/azureml-assets/azureml/assets/validate_assets.py
+++ b/scripts/azureml-assets/azureml/assets/validate_assets.py
@@ -79,8 +79,10 @@ def validate_assets(input_dirs: List[Path],
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-dirs", required=True, help="Comma-separated list of directories containing assets")
-    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME, help="Asset config file name to search for")
+    parser.add_argument("-i", "--input-dirs", required=True,
+                        help="Comma-separated list of directories containing assets")
+    parser.add_argument("-a", "--asset-config-filename", default=assets.DEFAULT_ASSET_FILENAME,
+                        help="Asset config file name to search for")
     args = parser.parse_args()
 
     # Convert comma-separated values to lists
diff --git a/scripts/azureml-assets/validation_rules.json b/scripts/azureml-assets/validation_rules.json
new file mode 100644
index 000000000..fc2610410
--- /dev/null
+++ b/scripts/azureml-assets/validation_rules.json
@@ -0,0 +1,24 @@
+{
+  "pep8": {
+    "ignore-file": [
+      "azureml/assets/__init__.py:F401",
+      "azureml/assets/environment/__init__.py:F401",
+      "azureml/assets/util/__init__.py:F401"
+    ],
+    "justifications": [
+      "__init__.py - namespace level files excluded from validation"
+    ]
+  },
+  "doc": {
+    "exclude": [
+      "azureml/assets/__init__.py",
+      "azureml/assets/environment/__init__.py",
+      "azureml/assets/util/__init__.py",
+      "."
+    ],
+    "justifications": [
+      "__init__.py - namespace level files excluded from validation",
+      ". - not enabled yet"
+    ]
+  }
+}
diff --git a/scripts/flake_rules.json b/scripts/flake_rules.json
deleted file mode 100644
index 065c489c9..000000000
--- a/scripts/flake_rules.json
+++ /dev/null
@@ -1,12 +0,0 @@
-{
-  "pep8": {
-    "ignore-file": [
-      "azureml-assets/azureml/assets/__init__.py:F401",
-      "azureml-assets/azureml/assets/environment/__init__.py:F401",
-      "azureml-assets/azureml/assets/util/__init__.py:F401"
-    ],
-    "justifications": [
-      "__init__.py - namespace level files excluded from validation"
-    ]
-  }
-}
\ No newline at end of file
diff --git a/scripts/script-check-requirements.txt b/scripts/script-check-requirements.txt
index 369eff651..54b8cabd6 100644
--- a/scripts/script-check-requirements.txt
+++ b/scripts/script-check-requirements.txt
@@ -1 +1,2 @@
-flake8==4.0.1
\ No newline at end of file
+flake8==4.0.1
+pydocstyle==6.1.1
\ No newline at end of file
diff --git a/scripts/validation/code_health.py b/scripts/validation/code_health.py
index 701c64c67..a5c4cfe2f 100644
--- a/scripts/validation/code_health.py
+++ b/scripts/validation/code_health.py
@@ -7,32 +7,74 @@ import json
 import sys
 from pathlib import Path
 from subprocess import run, PIPE, STDOUT
-from typing import Dict, List, Tuple
+from typing import Dict, List, Set, Tuple
 
-FLAKE_RULES_FILE = "flake_rules.json"
-DEFAULT_MAX_LINE_LENGTH = 10000
-IGNORE = "ignore"
-IGNORE_FILE = "ignore-file"
-EXCLUDE = "exclude"
-MAX_LINE_LENGTH = "max-line-length"
+RULES_FILENAME = "validation_rules.json"
 
 
-def run_flake8(testpath: Path, flake_rules: Dict[str, List[str]]) -> Tuple[int, str]:
-    ignore = flake_rules.get(IGNORE, [])
-    file_ignore = flake_rules.get(IGNORE_FILE, [])
-    exclude = flake_rules.get(EXCLUDE, [])
-    max_line_length = flake_rules.get(MAX_LINE_LENGTH, DEFAULT_MAX_LINE_LENGTH)
+class Rules:
+    ROOT_KEY = "pep8"
+    IGNORE = "ignore"
+    IGNORE_FILE = "ignore-file"
+    EXCLUDE = "exclude"
+    MAX_LINE_LENGTH = "max-line-length"
+    DEFAULT_MAX_LINE_LENGTH = 10000
+
+    def __init__(self, file_name: Path = None):
+        if file_name is not None and file_name.exists():
+            parent_path = file_name.parent
+            # Load rules from file
+            with open(file_name) as f:
+                rules = json.load(f).get(self.ROOT_KEY, {})
+                self.ignore = set(rules.get(self.IGNORE, []))
+                self.ignore_file = self._parse_ignore_file(parent_path, rules.get(self.IGNORE_FILE, []))
+                self.exclude = {parent_path / p for p in rules.get(self.EXCLUDE, [])}
+                self.max_line_length = rules.get(self.MAX_LINE_LENGTH)
+        else:
+            # Initialize empty
+            self.ignore = set()
+            self.ignore_file = {}
+            self.exclude = set()
+            self.max_line_length = None
+
+    @staticmethod
+    def _parse_ignore_file(parent_path: Path, ignore_file: List[str]) -> Dict[str, Set[str]]:
+        results = {}
+        for pair in ignore_file:
+            file, file_rules = pair.split(":")
+            file = parent_path / file
+            file_rules = {r.strip() for r in file_rules.split(",") if not r.isspace()}
+            results[file] = file_rules
+        return results
+
+    def get_effective_max_line_length(self) -> int:
+        return self.max_line_length or self.DEFAULT_MAX_LINE_LENGTH
+
+    def __or__(self, other: "Rules") -> "Rules":
+        rules = Rules()
+        rules.ignore = self.ignore | other.ignore
+        for key in self.ignore_file.keys() | other.ignore_file.keys():
+            rules.ignore_file[key] = self.ignore_file.get(key, set()) | other.ignore_file.get(key, set())
+        rules.exclude = self.exclude | other.exclude
+        rules.max_line_length = other.max_line_length or self.max_line_length
+        return rules
+
+
+def run_flake8(testpath: Path, rules: Rules) -> Tuple[int, str]:
     cmd = [
         "flake8",
-        f"--max-line-length={max_line_length}",
+        f"--max-line-length={rules.get_effective_max_line_length()}",
         str(testpath)
     ]
-    if exclude:
-        cmd.insert(1, "--exclude={}".format(",".join(exclude)))
-    if file_ignore:
-        cmd.insert(1, "--per-file-ignores={}".format(",".join(file_ignore)))
-    if ignore:
-        cmd.insert(1, "--ignore={}".format(",".join(ignore)))
+    if rules.exclude:
+        cmd.insert(1, "--exclude={}".format(",".join([str(e) for e in rules.exclude])))
+    if rules.ignore_file:
+        file_ignore_list = []
+        for file, ignores in rules.ignore_file.items():
+            file_ignore_list.extend([f"{file}:{i}" for i in ignores])
+        cmd.insert(1, "--per-file-ignores={}".format(",".join(file_ignore_list)))
+    if rules.ignore:
+        cmd.insert(1, "--ignore={}".format(",".join(rules.ignore)))
 
     print(f"Running {cmd}")
     p = run(cmd,
@@ -42,68 +84,35 @@ def run_flake8(testpath: Path, flake_rules: Dict[str, List[str]]) -> Tuple[int,
     return p.stdout.decode()
 
 
-def load_rules(testpath: Path) -> Dict[str, List[str]]:
-    flake_rules_file = testpath / FLAKE_RULES_FILE
-    flake_rules = {}
-    if flake_rules_file.exists():
-        with open(flake_rules_file) as f:
-            flake_rules = json.load(f).get('pep8', {})
-
-    # Handle relative paths
-    if IGNORE_FILE in flake_rules:
-        file_ignore = []
-        for pair in flake_rules[IGNORE_FILE]:
-            file, rules = pair.split(":")
-            file_resolved = str(testpath / file)
-            file_ignore.append(f"{file_resolved}:{rules}")
-        flake_rules[IGNORE_FILE] = file_ignore
-    if EXCLUDE in flake_rules:
-        flake_rules[EXCLUDE] = [str(testpath / p) for p in flake_rules[EXCLUDE]]
-
-    return flake_rules
-
-
-def combine_rules(rule_a: Dict[str, List[str]], rule_b: Dict[str, List[str]]) -> Dict[str, List[str]]:
-    rule = {}
-    rule[IGNORE] = rule_a.get(IGNORE, []) + rule_b.get(IGNORE, [])
-    rule[IGNORE_FILE] = rule_a.get(IGNORE_FILE, []) + rule_b.get(IGNORE_FILE, [])
-    rule[EXCLUDE] = rule_a.get(EXCLUDE, []) + rule_b.get(EXCLUDE, [])
-    rule[MAX_LINE_LENGTH] = min(rule_a.get(MAX_LINE_LENGTH, DEFAULT_MAX_LINE_LENGTH),
-                                rule_b.get(MAX_LINE_LENGTH, DEFAULT_MAX_LINE_LENGTH))
-
-    return rule
-
-
-def inherit_flake_rules(rootpath: Path, testpath: Path) -> Dict[str, List[str]]:
-    flake_rules = {}
-    upperpath = testpath
-    while upperpath != rootpath:
-        upperpath = upperpath.parent
-        flake_rules = combine_rules(flake_rules, load_rules(upperpath))
-    return flake_rules
+def inherit_rules(rootpath: Path, testpath: Path) -> Rules:
+    # Process paths from rootpath to testpath, to ensure max_line_length is calculated properly
+    paths = [p for p in testpath.parents if p == rootpath or p.is_relative_to(rootpath)]
+    paths.reverse()
+    rules = Rules()
+    for path in paths:
+        rules |= Rules(path / RULES_FILENAME)
+    return rules
 
 
 def test(rootpath: Path, testpath: Path) -> bool:
-    test_path_flake_rules = inherit_flake_rules(rootpath, testpath)
+    testpath_rules = inherit_rules(rootpath, testpath)
 
-    flake_rules_files = list(testpath.rglob(FLAKE_RULES_FILE))
-    dirs = [p.parent for p in flake_rules_files]
+    rules_files = list(testpath.rglob(RULES_FILENAME))
+    dirs = [p.parent for p in rules_files]
 
-    if not testpath == dirs:
+    if testpath not in dirs:
         dirs.insert(0, testpath)
 
-    output = []
+    errors = []
     for path in dirs:
-        flake_rules = {}
-        flake_rules[EXCLUDE] = [str(f.parent) for f in flake_rules_files if f.parent != path]
-        inherited_rules = inherit_flake_rules(testpath, path)
-        flake_rules = combine_rules(combine_rules(flake_rules, test_path_flake_rules),
-                                    combine_rules(inherited_rules, load_rules(path)))
-        output.extend([line for line in run_flake8(path, flake_rules).split("\n") if len(line) > 0])
+        rules = Rules()
+        rules.exclude = {d for d in dirs if d != path and not path.is_relative_to(d)}
+        rules |= testpath_rules | inherit_rules(testpath, path) | Rules(path / RULES_FILENAME)
+        errors.extend([line for line in run_flake8(path, rules).splitlines() if len(line) > 0])
 
-    if len(output) > 0:
+    if len(errors) > 0:
         print("flake8 errors:")
-        for line in output:
+        for line in errors:
             print(line)
         return False
     return True
@@ -112,8 +121,10 @@ def test(rootpath: Path, testpath: Path) -> bool:
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-directory", required=True, type=Path, help="Directory to validate")
-    parser.add_argument("-r", "--root-directory", type=Path, help="Root directory containing flake8 rules, must be a parent of --input-directory")
+    parser.add_argument("-i", "--input-directory", required=True, type=Path,
+                        help="Directory to validate")
+    parser.add_argument("-r", "--root-directory", type=Path,
+                        help="Root directory containing flake8 rules, must be a parent of --input-directory")
     args = parser.parse_args()
 
     # Handle directories
diff --git a/scripts/validation/copyright_validation.py b/scripts/validation/copyright_validation.py
index 3ce09366b..767e187eb 100644
--- a/scripts/validation/copyright_validation.py
+++ b/scripts/validation/copyright_validation.py
@@ -40,8 +40,10 @@ def test(testpaths: List[Path], excludes: List[Path] = []) -> bool:
 if __name__ == '__main__':
     # Handle command-line args
     parser = argparse.ArgumentParser()
-    parser.add_argument("-i", "--input-directories", required=True, type=Path, nargs='+', help="Directories to validate")
-    parser.add_argument("-e", "--excludes", default=[], type=Path, nargs='+', help="Directories to exclude")
+    parser.add_argument("-i", "--input-directories", required=True, type=Path, nargs='+',
+                        help="Directories to validate")
+    parser.add_argument("-e", "--excludes", default=[], type=Path, nargs='+',
+                        help="Directories to exclude")
     args = parser.parse_args()
 
     success = test(args.input_directories, args.excludes)
diff --git a/scripts/validation/doc_style.py b/scripts/validation/doc_style.py
new file mode 100644
index 000000000..1d4ea1239
--- /dev/null
+++ b/scripts/validation/doc_style.py
@@ -0,0 +1,149 @@
+# ---------------------------------------------------------
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# ---------------------------------------------------------
+
+import argparse
+import json
+import re
+import sys
+from pathlib import Path
+from subprocess import run, PIPE, STDOUT
+from typing import List, Set
+
+RULES_FILENAME = "validation_rules.json"
+FILE_NAME_PATTERN = re.compile(r"^(.+):\d+\s+")
+
+
+class Rules:
+    ROOT_KEY = "doc"
+    IGNORE = "ignore"
+    EXCLUDE = "exclude"
+    FORCE = "force"
+
+    def __init__(self, file_name: Path = None):
+        if file_name is not None and file_name.exists():
+            # Load rules from file
+            with open(file_name) as f:
+                rules = json.load(f).get(self.ROOT_KEY, {})
+                self.ignore = set(rules.get(self.IGNORE, []))
+                self.exclude = {file_name.parent / p for p in rules.get(self.EXCLUDE, [])}
+                self.force = set(rules.get(self.FORCE, []))
+        else:
+            # Initialize empty
+            self.ignore = set()
+            self.exclude = set()
+            self.force = set()
+
+    def __or__(self, other: "Rules") -> "Rules":
+        rules = Rules()
+        rules.ignore = self.ignore | other.ignore
+        rules.exclude = self.exclude | other.exclude
+        rules.force = self.force | other.force
+        return rules
+
+
+def run_docstyle(testpath: Path, rules: Rules):
+    cmd = [
+        "pydocstyle",
+        r"--match=.*\.py",
+        str(testpath)
+    ]
+
+    ignore = rules.ignore - rules.force
+    if ignore:
+        cmd.insert(1, "--ignore={}".format(",".join(ignore)))
+
+    print(f"Running {cmd}")
+    p = run(cmd,
+            stdout=PIPE,
+            stderr=STDOUT)
+
+    return p.stdout.decode()
+
+
+def filter_docstyle_output(output: str, rules: Rules) -> List[str]:
+    lines = [line for line in output.splitlines() if len(line) > 0]
+    if lines:
+        filtered_lines = []
+        for i in range(0, len(lines) - 1, 2):
+            line = lines[i]
+            match = FILE_NAME_PATTERN.match(line)
+            if not match:
+                raise Exception(f"Unable to extract filename from {line}")
+            file = Path(match.group(1))
+            file_is_excluded = False
+            for exclude in rules.exclude:
+                if file == exclude or (exclude.is_dir() and file.is_relative_to(exclude)):
+                    file_is_excluded = True
+                    break
+            if not file_is_excluded:
+                filtered_lines.append(line)
+                filtered_lines.append(lines[i + 1])
+        lines = filtered_lines
+
+    return lines
+
+
+def inherit_rules(rootpath: Path, testpath: Path) -> Rules:
+    rules = Rules()
+    upperpath = testpath
+    while upperpath != rootpath:
+        upperpath = upperpath.parent
+        rules |= Rules(upperpath / RULES_FILENAME)
+    return rules
+
+
+def test(rootpath: Path, testpath: Path, force: Set[str] = {}) -> bool:
+    testpath_rules = inherit_rules(rootpath, testpath)
+
+    rules_files = list(testpath.rglob(RULES_FILENAME))
+    dirs = [p.parent for p in rules_files]
+
+    if testpath not in dirs:
+        dirs.insert(0, testpath)
+
+    errors = []
+    for path in dirs:
+        rules = Rules()
+        rules.exclude = {d for d in dirs if d != path and not path.is_relative_to(d)}
+        rules.force = force
+        rules |= testpath_rules | inherit_rules(testpath, path) | Rules(path / RULES_FILENAME)
+
+        output = run_docstyle(path, rules)
+        filtered_output = filter_docstyle_output(output, rules)
+        errors.extend(filtered_output)
+
+    if len(errors) > 0:
+        print("pydocstyle errors:")
+        for line in errors:
+            print(line)
+        return False
+    return True
+
+
+if __name__ == '__main__':
+    # Handle command-line args
+    parser = argparse.ArgumentParser()
+    parser.add_argument("-i", "--input-directory", required=True, type=Path,
+                        help="Directory to validate")
+    parser.add_argument("-r", "--root-directory", type=Path,
+                        help="Root directory containing docstyle rules, must be a parent of --input-directory")
+    parser.add_argument("-f", "--force", default="",
+                        help="Comma-separated list of rules that can't be ignored")
+    args = parser.parse_args()
+
+    # Handle directories
+    input_directory = args.input_directory
+    root_directory = args.root_directory
+    if root_directory is None:
+        root_directory = args.input_directory
+    elif not input_directory.is_relative_to(root_directory):
+        parser.error(f"{root_directory} is not a parent directory of {input_directory}")
+
+    # Parse forced rules
+    force = {r.strip() for r in args.force.split(",") if not r.isspace()}
+
+    success = test(root_directory, input_directory, force)
+
+    if not success:
+        sys.exit(1)
diff --git a/scripts/validation/validation_rules.json b/scripts/validation/validation_rules.json
new file mode 100644
index 000000000..c63b2b207
--- /dev/null
+++ b/scripts/validation/validation_rules.json
@@ -0,0 +1,10 @@
+{
+  "doc": {
+    "exclude": [
+      "."
+    ],
+    "justifications": [
+      ". - not enabled yet"
+    ]
+  }
+}
diff --git a/test/test_update_assets.py b/test/test_update_assets.py
index c309ad32b..b50231b43 100644
--- a/test/test_update_assets.py
+++ b/test/test_update_assets.py
@@ -31,7 +31,9 @@ def test_validate_assets(test_subdir: str, create_tag: bool):
     expected_dir = test_dir / "expected"
 
     # Temp directory helps keep the original release directory clean
-    with tempfile.TemporaryDirectory(prefix="release-") as temp_dir1, tempfile.TemporaryDirectory(prefix="output-") as temp_dir2, tempfile.TemporaryDirectory(prefix="expected-") as temp_dir3:
+    with tempfile.TemporaryDirectory(prefix="release-") as temp_dir1, \
+            tempfile.TemporaryDirectory(prefix="output-") as temp_dir2, \
+            tempfile.TemporaryDirectory(prefix="expected-") as temp_dir3:
         temp_release_path = Path(temp_dir1)
         temp_output_path = Path(temp_dir2)
         temp_expected_path = Path(temp_dir3)
diff --git a/test/validation_rules.json b/test/validation_rules.json
new file mode 100644
index 000000000..c63b2b207
--- /dev/null
+++ b/test/validation_rules.json
@@ -0,0 +1,10 @@
+{
+  "doc": {
+    "exclude": [
+      "."
+    ],
+    "justifications": [
+      ". - not enabled yet"
+    ]
+  }
+}
diff --git a/training/validation_rules.json b/training/validation_rules.json
new file mode 100644
index 000000000..c63b2b207
--- /dev/null
+++ b/training/validation_rules.json
@@ -0,0 +1,10 @@
+{
+  "doc": {
+    "exclude": [
+      "."
+    ],
+    "justifications": [
+      ". - not enabled yet"
+    ]
+  }
+}
diff --git a/validation_rules.json b/validation_rules.json
new file mode 100644
index 000000000..32e643497
--- /dev/null
+++ b/validation_rules.json
@@ -0,0 +1,5 @@
+{
+  "pep8": {
+    "max-line-length": 119
+  }
+}