rasa-pro 3.13.7__tar.gz → 3.13.9__tar.gz

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rasa-pro
-Version: 3.13.7
+Version: 3.13.9
 Summary: State-of-the-art open-core Conversational AI framework for Enterprises that natively leverages generative AI for effortless assistant development.
 Keywords: nlp,machine-learning,machine-learning-library,bot,bots,botkit,rasa conversational-agents,conversational-ai,chatbot,chatbot-framework,bot-framework
 Author: Rasa Technologies GmbH
@@ -102,7 +102,7 @@ Requires-Dist: randomname (>=0.2.1,<0.3.0)
 Requires-Dist: rasa-sdk (==3.13.0)
 Requires-Dist: redis (>=4.6.0,<6.0)
 Requires-Dist: regex (>=2024.7.24,<2024.8.0)
-Requires-Dist: requests (>=2.32.3,<2.33.0)
+Requires-Dist: requests (>=2.32.5,<2.33.0)
 Requires-Dist: rich (>=13.4.2,<14.0.0)
 Requires-Dist: rocketchat_API (>=1.32.0,<1.33.0)
 Requires-Dist: ruamel.yaml (>=0.17.21,<0.17.22)
@@ -117,7 +117,7 @@ Requires-Dist: sentencepiece[sentencepiece] (>=0.1.99,<0.2.0) ; extra == "transf
 Requires-Dist: sentry-sdk (>=2.8.0,<3)
 Requires-Dist: setuptools (>=78.1.1,<78.2.0)
 Requires-Dist: sklearn-crfsuite (>=0.3.6,<0.4.0)
-Requires-Dist: skops (>=0.10.0,<0.11.0)
+Requires-Dist: skops (>=0.13.0,<0.14.0)
 Requires-Dist: slack-sdk (>=3.27.1,<3.28.0)
 Requires-Dist: spacy (>=3.5.4,<4.0.0) ; extra == "spacy" or extra == "full"
 Requires-Dist: structlog (>=23.1.0,<23.2.0)

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "rasa-pro"
-version = "3.13.7"
+version = "3.13.9"
 description = "State-of-the-art open-core Conversational AI framework for Enterprises that natively leverages generative AI for effortless assistant development."
 authors = [ "Rasa Technologies GmbH <hi@rasa.com>",]
 maintainers = [ "Tom Bocklisch <tom@rasa.com>",]
@@ -127,7 +127,7 @@ python = ">=3.9.2,<3.12"
 boto3 = "~1.35.15"
 google-cloud-storage = "~2.14.0"
 azure-storage-blob = ">=12.16.0, <12.17.0"
-requests = "~2.32.3"
+requests = "~2.32.5"
 matplotlib = ">=3.7,<3.8"
 attrs = ">=23.1,<23.2"
 jsonpickle = ">=3.3.0,<3.4"
@@ -218,7 +218,7 @@ types-protobuf = "4.25.0.20240417"
 litellm = "~1.69.0"
 openpyxl = "^3.1.5"
 diskcache = "~5.6.3"
-skops = "~0.10.0"
+skops = "~0.13.0"
 safetensors = "~0.4.5"
 dask = "~2024.7.0" # later versions drop python 3.9 support
 numpy = "~1.26.4"
@@ -384,6 +384,8 @@ convention = "google"
 
 
 
+
+
 [tool.poetry.group.dev.dependencies]
 ruff = ">=0.3.5,<0.4.0"
 docker = "~7.1.0"

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/dialogue_understanding/generator/flow_retrieval.py

@@ -249,6 +249,16 @@ class FlowRetrieval(EmbeddingsHealthCheckMixin):
         )
 
         flows_to_embedd = flows.exclude_link_only_flows()
+
+        if not flows_to_embedd:
+            structlogger.debug(
+                "flow_retrieval.populate_vector_store.no_flows_to_embed",
+                event_info=(
+                    "No flows to embed in the vector store, skipping population."
+                ),
+            )
+            return
+
         embeddings = self._create_embedder(self.config)
         documents = self._generate_flow_documents(flows_to_embedd, domain)
         try:

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/engine/storage/local_model_storage.py

@@ -1,14 +1,16 @@
 from __future__ import annotations
 
 import logging
+import os
 import shutil
 import sys
+import tarfile
 import tempfile
 import uuid
 from contextlib import contextmanager
 from datetime import datetime
 from pathlib import Path
-from typing import Generator, Optional, Text, Tuple, Union
+from typing import Callable, Generator, List, Optional, Text, Tuple, Union
 
 from tarsafe import TarSafe
 
@@ -57,6 +59,35 @@ def windows_safe_temporary_directory(
             yield temporary_directory
 
 
+def filter_normpath(member: tarfile.TarInfo, dest_path: str) -> tarfile.TarInfo:
+    """Normalize tar member paths for safe extraction"""
+    if member.name:
+        member.name = os.path.normpath(member.name)
+    return member
+
+
+FilterFunction = Callable[[tarfile.TarInfo, str], Optional[tarfile.TarInfo]]
+
+
+def create_combined_filter(existing_filter: Optional[FilterFunction]) -> FilterFunction:
+    """Create a filter that combines existing filter with path normalization"""
+
+    def combined_filter(
+        member: tarfile.TarInfo, dest_path: str
+    ) -> Optional[tarfile.TarInfo]:
+        """Apply existing filter first, then path normalization"""
+        if existing_filter is not None:
+            filtered_member = existing_filter(member, dest_path)
+            if filtered_member is None:
+                return None  # Rejected by existing filter
+            member = filtered_member  # Use the filtered result
+
+        # Apply our path normalization
+        return filter_normpath(member, dest_path)
+
+    return combined_filter
+
+
 class LocalModelStorage(ModelStorage):
     """Stores and provides output of `GraphComponents` on local disk."""
 
@@ -122,9 +153,29 @@ class LocalModelStorage(ModelStorage):
                 # this restriction in environments where it's not possible
                 # to override this behavior, mostly for internal policy reasons
                 # reference: https://stackoverflow.com/a/49102229
-                tar.extractall(f"\\\\?\\{temporary_directory}")
+                try:
+                    # Use extraction filter to normalize paths for compatibility
+                    # before trying the \\?\ prefix approach first
+                    prev_filter = getattr(tar, "extraction_filter", None)
+                    tar.extraction_filter = create_combined_filter(prev_filter)
+                    tar.extractall(
+                        f"\\\\?\\{temporary_directory}",
+                        members=yield_safe_members(tar.getmembers()),
+                    )
+                except Exception:
+                    # Fallback for Python versions with tarfile security fix
+                    logger.warning(
+                        "Failed to extract model archive with long path support. "
+                        "Falling back to regular extraction."
+                    )
+                    tar.extractall(
+                        temporary_directory,
+                        members=yield_safe_members(tar.getmembers()),
+                    )
             else:
-                tar.extractall(temporary_directory)
+                tar.extractall(
+                    temporary_directory, members=yield_safe_members(tar.getmembers())
+                )
         LocalModelStorage._assert_not_rasa2_archive(temporary_directory)
 
     @staticmethod
@@ -244,3 +295,32 @@ class LocalModelStorage(ModelStorage):
             core_target=model_configuration.core_target,
             nlu_target=model_configuration.nlu_target,
         )
+
+
+def yield_safe_members(
+    members: List[tarfile.TarInfo],
+) -> Generator[tarfile.TarInfo, None, None]:
+    """
+    Filter function for tar.extractall members parameter.
+    Validates each member and yields only safe ones.
+
+    Args:
+        members: Iterator of TarInfo objects from tar.getmembers()
+
+    Yields:
+        TarInfo: Safe members to extract
+    """
+    for member in members:
+        # Skip absolute paths
+        if Path(member.name).is_absolute():
+            continue
+
+        # Skip paths with directory traversal sequences
+        if ".." in member.name or "\\.." in member.name:
+            continue
+
+        # Skip special file types unless you need them
+        if member.isdev() or member.issym():
+            continue
+
+        yield member

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/shared/providers/_utils.py

@@ -1,87 +1,103 @@
 from typing import Any, Dict, Optional
 
+import boto3
 import structlog
-from litellm import validate_environment
+from botocore.exceptions import BotoCoreError, ClientError
 
 from rasa.shared.constants import (
     API_BASE_CONFIG_KEY,
     API_VERSION_CONFIG_KEY,
     AWS_ACCESS_KEY_ID_CONFIG_KEY,
-    AWS_ACCESS_KEY_ID_ENV_VAR,
+    AWS_BEDROCK_PROVIDER,
     AWS_REGION_NAME_CONFIG_KEY,
-    AWS_REGION_NAME_ENV_VAR,
+    AWS_SAGEMAKER_CHAT_PROVIDER,
+    AWS_SAGEMAKER_PROVIDER,
     AWS_SECRET_ACCESS_KEY_CONFIG_KEY,
-    AWS_SECRET_ACCESS_KEY_ENV_VAR,
     AWS_SESSION_TOKEN_CONFIG_KEY,
-    AWS_SESSION_TOKEN_ENV_VAR,
     AZURE_API_BASE_ENV_VAR,
     AZURE_API_VERSION_ENV_VAR,
     DEPLOYMENT_CONFIG_KEY,
 )
 from rasa.shared.exceptions import ProviderClientValidationError
-from rasa.shared.providers.embedding._base_litellm_embedding_client import (
-    _VALIDATE_ENVIRONMENT_MISSING_KEYS_KEY,
-)
+from rasa.shared.utils.io import resolve_environment_variables
 
 structlogger = structlog.get_logger()
 
 
 def validate_aws_setup_for_litellm_clients(
-    litellm_model_name: str, litellm_call_kwargs: dict, source_log: str
+    litellm_model_name: str, litellm_call_kwargs: Dict, source_log: str, provider: str
 ) -> None:
-    """Validates the AWS setup for LiteLLM clients to ensure all required
-    environment variables or corresponding call kwargs are set.
+    """Validates the AWS setup for LiteLLM clients to ensure credentials are set.
 
     Args:
         litellm_model_name (str): The name of the LiteLLM model being validated.
         litellm_call_kwargs (dict): Additional keyword arguments passed to the client,
             which may include configuration values for AWS credentials.
         source_log (str): The source log identifier for structured logging.
+        provider (str): The provider for which the validation is being performed.
 
     Raises:
         ProviderClientValidationError: If any required AWS environment variable
             or corresponding configuration key is missing.
     """
-
-    # Mapping of environment variable names to their corresponding config keys
-    envs_to_args = {
-        AWS_ACCESS_KEY_ID_ENV_VAR: AWS_ACCESS_KEY_ID_CONFIG_KEY,
-        AWS_SECRET_ACCESS_KEY_ENV_VAR: AWS_SECRET_ACCESS_KEY_CONFIG_KEY,
-        AWS_REGION_NAME_ENV_VAR: AWS_REGION_NAME_CONFIG_KEY,
-        AWS_SESSION_TOKEN_ENV_VAR: AWS_SESSION_TOKEN_CONFIG_KEY,
-    }
-
-    # Validate the environment setup for the model
-    validation_info = validate_environment(litellm_model_name)
-    missing_environment_variables = validation_info.get(
-        _VALIDATE_ENVIRONMENT_MISSING_KEYS_KEY, []
+    # expand environment variables if referenced in the config
+    resolved_litellm_call_kwargs: Dict = resolve_environment_variables(
+        litellm_call_kwargs
+    )  # type: ignore[assignment]
+
+    # boto3 only accepts bedrock and sagemaker as valid clients
+    # therefore we need to convert the provider name if it is defined
+    # as sagemaker_chat
+    provider = (
+        AWS_SAGEMAKER_PROVIDER if provider == AWS_SAGEMAKER_CHAT_PROVIDER else provider
     )
-    # Filter out missing environment variables that have been set trough arguments
-    # in extra parameters
-    missing_environment_variables = [
-        missing_env_var
-        for missing_env_var in missing_environment_variables
-        if litellm_call_kwargs.get(envs_to_args.get(missing_env_var)) is None
-    ]
 
-    if missing_environment_variables:
-        missing_environment_details = [
-            (
-                f"'{missing_env_var}' environment variable or "
-                f"'{envs_to_args.get(missing_env_var)}' config key"
-            )
-            for missing_env_var in missing_environment_variables
+    # if the AWS credentials are defined in the endpoints yaml model config,
+    # either as referenced secret env vars or direct values, we need to pass them
+    # to the boto3 client to ensure that the client can connect to the AWS service.
+    additional_kwargs: Dict[str, Any] = {}
+    if AWS_ACCESS_KEY_ID_CONFIG_KEY in resolved_litellm_call_kwargs:
+        additional_kwargs[AWS_ACCESS_KEY_ID_CONFIG_KEY] = resolved_litellm_call_kwargs[
+            AWS_ACCESS_KEY_ID_CONFIG_KEY
         ]
+    if AWS_SECRET_ACCESS_KEY_CONFIG_KEY in resolved_litellm_call_kwargs:
+        additional_kwargs[AWS_SECRET_ACCESS_KEY_CONFIG_KEY] = (
+            resolved_litellm_call_kwargs[AWS_SECRET_ACCESS_KEY_CONFIG_KEY]
+        )
+    if AWS_SESSION_TOKEN_CONFIG_KEY in resolved_litellm_call_kwargs:
+        additional_kwargs[AWS_SESSION_TOKEN_CONFIG_KEY] = resolved_litellm_call_kwargs[
+            AWS_SESSION_TOKEN_CONFIG_KEY
+        ]
+    if AWS_REGION_NAME_CONFIG_KEY in resolved_litellm_call_kwargs:
+        additional_kwargs["region_name"] = resolved_litellm_call_kwargs[
+            AWS_REGION_NAME_CONFIG_KEY
+        ]
+
+    try:
+        # We are using the boto3 client because it can discover the AWS credentials
+        # from the environment variables, credentials file, or IAM roles.
+        # This is necessary to ensure that the client can connect to the AWS service.
+        aws_client = boto3.client(provider, **additional_kwargs)
+
+        # Using different method calls available to different AWS clients
+        # to test the connection
+        if provider == AWS_SAGEMAKER_PROVIDER:
+            aws_client.list_models()
+        elif provider == AWS_BEDROCK_PROVIDER:
+            aws_client.get_model_invocation_logging_configuration()
+
+    except (ClientError, BotoCoreError) as exc:
         event_info = (
-            f"The following environment variables or configuration keys are "
-            f"missing: "
-            f"{', '.join(missing_environment_details)}. "
-            f"These settings are required for API calls."
+            f"Failed to validate AWS setup for LiteLLM clients: {exc}. "
+            f"Ensure that you are using one of the available authentication methods:"
+            f"credentials file, environment variables, or IAM roles. "
+            f"Also, ensure that the AWS region is set correctly. "
         )
         structlogger.error(
-            f"{source_log}.validate_aws_environment_variables",
+            f"{source_log}.validate_aws_credentials_for_litellm_clients",
             event_info=event_info,
-            missing_environment_variables=missing_environment_variables,
+            exception=str(exc),
+            model_name=litellm_model_name,
         )
         raise ProviderClientValidationError(event_info)
 

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/shared/providers/embedding/default_litellm_embedding_client.py

@@ -37,6 +37,7 @@ class DefaultLiteLLMEmbeddingClient(_BaseLiteLLMEmbeddingClient):
 
     @classmethod
     def from_config(cls, config: Dict[str, Any]) -> "DefaultLiteLLMEmbeddingClient":
+        """Creates a DefaultLiteLLMEmbeddingClient instance from a config dict."""
         default_config = DefaultLiteLLMClientConfig.from_dict(config)
         return cls(
             model=default_config.model,
@@ -121,6 +122,7 @@ class DefaultLiteLLMEmbeddingClient(_BaseLiteLLMEmbeddingClient):
                 self._litellm_model_name,
                 self._litellm_extra_parameters,
                 "default_litellm_embedding_client",
+                provider=self.provider,
             )
         else:
             super().validate_client_setup()

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/shared/providers/llm/default_litellm_llm_client.py

@@ -39,6 +39,7 @@ class DefaultLiteLLMClient(_BaseLiteLLMClient):
 
     @classmethod
     def from_config(cls, config: Dict[str, Any]) -> DefaultLiteLLMClient:
+        """Creates a DefaultLiteLLMClient instance from a configuration dictionary."""
         default_config = DefaultLiteLLMClientConfig.from_dict(config)
         return cls(
             model=default_config.model,
@@ -110,6 +111,7 @@ class DefaultLiteLLMClient(_BaseLiteLLMClient):
                 self._litellm_model_name,
                 self._litellm_extra_parameters,
                 "default_litellm_llm_client",
+                provider=self.provider,
             )
         else:
             super().validate_client_setup()

{rasa_pro-3.13.7 → rasa_pro-3.13.9}/rasa/version.py

@@ -1,3 +1,3 @@
 # this file will automatically be changed,
 # do not add anything but the version number here!
-__version__ = "3.13.7"
+__version__ = "3.13.9"