rasa-pro 3.15.0a1__tar.gz → 3.15.0a3__tar.gz

{rasa_pro-3.15.0a1 → rasa_pro-3.15.0a3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rasa-pro
-Version: 3.15.0a1
+Version: 3.15.0a3
 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
@@ -39,16 +39,16 @@ Requires-Dist: apscheduler (>=3.11,<3.12)
 Requires-Dist: attrs (>=23.1,<25.0)
 Requires-Dist: audioop-lts (>=0.2.2,<0.3.0) ; python_version >= "3.13"
 Requires-Dist: aws-msk-iam-sasl-signer-python (>=1.0.2,<1.1.0)
-Requires-Dist: azure-identity (>=1.24.0,<1.25.0)
-Requires-Dist: azure-storage-blob (>=12.26.0,<12.27.0)
-Requires-Dist: boto3 (>=1.40.21,<1.41.0)
+Requires-Dist: azure-identity (>=1.25.1,<1.26.0)
+Requires-Dist: azure-storage-blob (>=12.27.0,<12.28.0)
+Requires-Dist: boto3 (>=1.40.60,<1.41.0)
 Requires-Dist: certifi (>=2025.10.5,<2025.11.0)
 Requires-Dist: colorama (>=0.4.6,<0.5.0) ; sys_platform == "win32"
 Requires-Dist: colorclass (>=2.2,<2.3)
 Requires-Dist: coloredlogs (>=15,<16)
 Requires-Dist: colorhash (>=2.0,<2.1.0)
 Requires-Dist: confluent-kafka (>=2.11.0,<3.0.0)
-Requires-Dist: cryptography (>=45.0.6,<45.1.0)
+Requires-Dist: cryptography (>=46.0.3,<46.1.0)
 Requires-Dist: cvg-python-sdk (>=0.5.1,<0.6.0) ; extra == "full" or extra == "channels"
 Requires-Dist: dask (>=2024.8.0,<2024.9.0)
 Requires-Dist: demoji (>=1.1.0,<2.0.0)
@@ -59,8 +59,8 @@ Requires-Dist: fbmessenger (>=6.0.0,<6.1.0) ; extra == "full" or extra == "chann
 Requires-Dist: github3.py (>=3.2.0,<3.3.0) ; extra == "gh-release-notes"
 Requires-Dist: gitpython (>=3.1.41,<3.2.0) ; extra == "full"
 Requires-Dist: gliner (>=0.2.20,<0.3.0) ; extra == "full" or extra == "pii"
-Requires-Dist: google-auth (>=2.40.3,<2.41.0)
-Requires-Dist: google-cloud-storage (>=2.14.0,<2.15.0)
+Requires-Dist: google-auth (>=2.41.1,<2.42.0)
+Requires-Dist: google-cloud-storage (>=3.4.1,<3.5.0)
 Requires-Dist: hvac (>=2.3.0,<2.4.0)
 Requires-Dist: importlib-metadata (>=8.5.0,<8.6.0)
 Requires-Dist: importlib-resources (>=6.5.2,<7.0.0)
@@ -104,7 +104,7 @@ Requires-Dist: python-dateutil (>=2.8.2,<2.9.0)
 Requires-Dist: python-dotenv (>=1.0.1,<2.0.0)
 Requires-Dist: python-engineio (>=4.12.2,<4.13.0)
 Requires-Dist: python-keycloak (>=5.8.1,<5.9.0)
-Requires-Dist: python-socketio (>=5.13,<6)
+Requires-Dist: python-socketio (>=5.14.0,<5.15.0)
 Requires-Dist: pytz (>=2022.7.1,<2023.0)
 Requires-Dist: pyyaml (>=6.0.2,<6.1.0)
 Requires-Dist: qdrant-client (>=1.9.1,<1.10.0)
@@ -127,14 +127,14 @@ Requires-Dist: scikit-learn (>=1.6.1,<1.7.0)
 Requires-Dist: scipy (>=1.13.1,<1.14.0) ; python_version < "3.12"
 Requires-Dist: scipy (>=1.14.0,<1.15.0) ; python_version >= "3.12"
 Requires-Dist: sentencepiece (>=0.1.99,<0.2.0) ; (python_version < "3.12") and (extra == "transformers" or extra == "full" or extra == "full" or extra == "nlu")
-Requires-Dist: sentry-sdk (>=2.8.0,<3)
+Requires-Dist: sentry-sdk (>=2.35.2,<2.36.0)
 Requires-Dist: setuptools (>=78.1.1,<78.2.0)
 Requires-Dist: sklearn-crfsuite (>=0.5.0,<0.6.0) ; extra == "full" or extra == "nlu"
 Requires-Dist: skops (>=0.13.0,<0.14.0) ; extra == "full" or extra == "nlu"
-Requires-Dist: slack-sdk (>=3.36.0,<3.37.0) ; extra == "full" or extra == "channels"
+Requires-Dist: slack-sdk (>=3.37.0,<3.38.0) ; extra == "full" or extra == "channels"
 Requires-Dist: spacy (>=3.5.4,<4.0.0) ; extra == "spacy" or extra == "full" or extra == "nlu"
 Requires-Dist: structlog (>=25.4.0,<25.5.0)
-Requires-Dist: structlog-sentry (>=2.0.3,<2.1.0)
+Requires-Dist: structlog-sentry (>=2.2.1,<2.3.0)
 Requires-Dist: tarsafe (>=0.0.5,<0.0.6)
 Requires-Dist: tenacity (>=8.4.1,<8.5.0)
 Requires-Dist: tensorflow (>=2.19.0,<3.0.0) ; (sys_platform != "darwin" or platform_machine != "arm64" and python_version < "3.12") and (extra == "full" or extra == "nlu")
@@ -149,7 +149,7 @@ Requires-Dist: tf-keras (>=2.15.0,<3.0.0) ; (python_version < "3.12") and (extra
 Requires-Dist: tiktoken (>=0.9.0,<0.10.0)
 Requires-Dist: tqdm (>=4.66.2,<5.0.0)
 Requires-Dist: transformers (>=4.38.2,<4.39.0) ; extra == "transformers" or extra == "full" or extra == "nlu"
-Requires-Dist: twilio (>=9.7.2,<9.8.0) ; extra == "full" or extra == "channels"
+Requires-Dist: twilio (>=9.8.4,<9.9.0) ; extra == "full" or extra == "channels"
 Requires-Dist: types-cachetools (>=6.2.0.20250827,<6.3.0.0)
 Requires-Dist: types-protobuf (==4.25.0.20240417)
 Requires-Dist: typing-extensions (>=4.7.1,<5.0.0)

{rasa_pro-3.15.0a1 → rasa_pro-3.15.0a3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "rasa-pro"
-version = "3.15.0a1"
+version = "3.15.0a3"
 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>",]
@@ -140,9 +140,9 @@ exclude = [ "eggs", ".git", ".pytest_cache", "build", "dist",]
 
 [tool.poetry.dependencies]
 python = ">=3.10.0,<3.14"
-boto3 = "~1.40.21"
-google-cloud-storage = "~2.14.0"
-azure-storage-blob = "~12.26.0"
+boto3 = "~1.40.60"
+google-cloud-storage = "~3.4.1"
+azure-storage-blob = "~12.27.0"
 requests = "~2.32.5"
 matplotlib = "~3.9.4"
 attrs = ">=23.1,<25.0"
@@ -170,7 +170,7 @@ websockets = ">=10.4,<11.0"
 aiohttp = ">=3.10,<3.11"
 questionary = ">=2.1.1,<2.2.0"
 prompt-toolkit = "^3.0.28,<3.0.29"
-python-socketio = ">=5.13,<6"
+python-socketio = "~5.14.0"
 python-engineio = "~4.12.2" # skip 5.0.0 upgrade
 pydot = "~1.4"
 SQLAlchemy = ">=2.0.42,<2.1.0"
@@ -180,12 +180,12 @@ protobuf = "~5.29.5"
 setuptools = "~78.1.1"
 ujson = ">=5.8,<6.0"
 regex = "~2024.7.24" # required by litellm + whitespacetokenizer
-sentry-sdk = ">=2.8.0,<3"
+sentry-sdk = "~2.35.2"
 aio-pika = ">=8.2.3,<9.4.4"
 typing-extensions = ">=4.7.1,<5.0.0"
 typing-utils = "^0.1.0"
 tarsafe = ">=0.0.5,<0.0.6"
-google-auth = "~2.40.3"
+google-auth = "~2.41.1"
 CacheControl = "~0.14.2"
 randomname = ">=0.2.1,<0.3.0"
 pluggy = "^1.2.0"
@@ -193,7 +193,7 @@ confluent-kafka = ">=2.11.0,<3.0.0"
 portalocker = "^2.7.0"
 pypred = "^0.4.0"
 structlog = ">=25.4.0,<25.5.0"
-structlog-sentry = ">=2.0.3,<2.1.0"
+structlog-sentry = "~2.2.1"
 langchain = "~0.3.27"
 langchain-community = "~0.3.29"
 jinja2 = "~3.1.6"
@@ -201,7 +201,7 @@ openai = "~1.68.2"
 dnspython = "2.6.1"
 wheel = ">=0.40.0"
 certifi = ">=2025.10.5,<2025.11.0"
-cryptography = "~45.0.6"
+cryptography = "~46.0.3"
 jsonpatch = "^1.33"
 importlib-resources = "^6.5.2"
 importlib-metadata = "~8.5.0"
@@ -227,7 +227,7 @@ dask = "~2024.8.0"  # later versions drop python 3.9 support
 numpy = "~2.1.3"  # langchain-community (>=0.3.29,<0.4.0) requires numpy (>=2.1.0) for python_version>='3.13'
 scikit-learn = "~1.6.1"
 demoji = "^1.1.0"  # required for voice and whitespacetokenizer
-azure-identity = "~1.24.0"
+azure-identity = "~1.25.1"
 langcodes = "^3.5.0"
 sanic-openapi = "^21.12.0"
 aioshutil = "~1.5"
@@ -297,7 +297,7 @@ version = "~6.0.0"
 optional = true
 
 [tool.poetry.dependencies.twilio]
-version = "~9.7.2"
+version = "~9.8.4"
 optional = true
 
 [tool.poetry.dependencies.webexteamssdk]
@@ -317,7 +317,7 @@ version = "~3.22.0"
 optional = true
 
 [tool.poetry.dependencies.slack-sdk]
-version = "~3.36.0"
+version = "~3.37.0"
 optional = true
 
 [tool.poetry.dependencies.cvg-python-sdk]
@@ -529,16 +529,6 @@ convention = "google"
 "pypred.Predicate".msg = "Only use the patched pypred module from rasa.utils.pypred."
 
 
-
-
-
-
-
-
-
-
-
-
 [tool.poetry.group.dev.dependencies]
 ruff = ">=0.3.5,<0.4.0"
 docker = "~7.1.0"

rasa_pro-3.15.0a3/rasa/builder/constants.py

@@ -0,0 +1,5 @@
+# Security limits for backup archive extraction
+MAX_BACKUP_SIZE = 100 * 1024 * 1024  # 100MB
+MAX_ARCHIVE_FILE_SIZE = 100 * 1024 * 1024  # 100MB
+MAX_ARCHIVE_TOTAL_SIZE = 500 * 1024 * 1024  # 500MB
+MAX_ARCHIVE_FILES = 10000

{rasa_pro-3.15.0a1 → rasa_pro-3.15.0a3}/rasa/builder/copilot/models.py

@@ -344,6 +344,55 @@ ChatMessage = Union[
 ]
 
 
+def create_chat_message_from_dict(message_data: Dict[str, Any]) -> ChatMessage:
+    """Parse a single chat message dictionary into a ChatMessage object.
+
+    This utility function manually parses a chat message dictionary into the
+    appropriate ChatMessage type based on its role field.
+
+    Args:
+        message_data: Dictionary containing chat message data
+
+    Returns:
+        Parsed ChatMessage object
+
+    Raises:
+        ValueError: If an unknown role is encountered
+
+    Example:
+        >>> message_data = {
+        ...     "role": "user",
+        ...     "content": [{"type": "text", "text": "Hello"}]
+        ... }
+        >>> message = parse_chat_message_from_dict(message_data)
+        >>> isinstance(message, UserChatMessage)
+        True
+        >>> message.role
+        'user'
+    """
+    available_roles = [ROLE_USER, ROLE_COPILOT, ROLE_COPILOT_INTERNAL]
+    role = message_data.get("role")
+
+    if role == ROLE_USER:
+        return UserChatMessage(**message_data)
+    elif role == ROLE_COPILOT:
+        return CopilotChatMessage(**message_data)
+    elif role == ROLE_COPILOT_INTERNAL:
+        return InternalCopilotRequestChatMessage(**message_data)
+    else:
+        message = (
+            f"Unknown role '{role}' in chat message. "
+            f"Available roles are: {', '.join(available_roles)}."
+        )
+        structlogger.error(
+            "models.create_chat_message_from_dict.unknown_role",
+            event_info=message,
+            role=role,
+            available_roles=available_roles,
+        )
+        raise ValueError(message)
+
+
 class CopilotContext(BaseModel):
     """Model containing the context used by the copilot to generate a response."""
 
@@ -391,37 +440,40 @@ class CopilotRequest(BaseModel):
 
     @field_validator("copilot_chat_history", mode="before")
     @classmethod
-    def parse_chat_history(cls, v: List[Dict[str, Any]]) -> List[ChatMessage]:
+    def parse_chat_history(
+        cls, v: Union[List[Dict[str, Any]], List[ChatMessage]]
+    ) -> List[ChatMessage]:
         """Manually parse chat history messages based on role field."""
+        # If already parsed ChatMessage objects, return them as-is
+        if (
+            v
+            and isinstance(v, list)
+            and all(isinstance(item, ChatMessage) for item in v)
+        ):
+            return v  # type: ignore[return-value]
+
+        # Check for mixed types (some ChatMessage, some not)
+        if (
+            v
+            and isinstance(v, list)
+            and any(isinstance(item, ChatMessage) for item in v)
+        ):
+            message = (
+                "Mixed types in copilot_chat_history: cannot mix ChatMessage objects"
+                "with other types."
+            )
+            structlog.get_logger().error(
+                "copilot_request.parse_chat_history.mixed_types",
+                event_info=message,
+                chat_history_types=[type(item) for item in v],
+            )
+            raise ValueError(message)
+
+        # Otherwise, parse from dictionaries
         parsed_messages: List[ChatMessage] = []
-        available_roles = [ROLE_USER, ROLE_COPILOT, ROLE_COPILOT_INTERNAL]
         for message_data in v:
-            role = message_data.get("role")
-
-            if role == ROLE_USER:
-                parsed_messages.append(UserChatMessage(**message_data))
-
-            elif role == ROLE_COPILOT:
-                parsed_messages.append(CopilotChatMessage(**message_data))
-
-            elif role == ROLE_COPILOT_INTERNAL:
-                parsed_messages.append(
-                    InternalCopilotRequestChatMessage(**message_data)
-                )
-
-            else:
-                message = (
-                    f"Unknown role '{role}' in chat message. "
-                    f"Available roles are: {', '.join(available_roles)}."
-                )
-                structlogger.error(
-                    "copilot_request.parse_chat_history.unknown_role",
-                    event_info=message,
-                    role=role,
-                    available_roles=available_roles,
-                )
-                raise ValueError(message)
-
+            chat_message = create_chat_message_from_dict(message_data)
+            parsed_messages.append(chat_message)
         return parsed_messages
 
     @property

{rasa_pro-3.15.0a1 → rasa_pro-3.15.0a3}/rasa/builder/download.py

@@ -1,11 +1,21 @@
 """Download utilities for bot projects."""
 
+import asyncio
 import io
 import os
 import sys
 import tarfile
+import tempfile
+from pathlib import Path
 from textwrap import dedent
 from typing import Dict, Optional
+from urllib.parse import urlparse
+
+import aiofiles
+import aiohttp
+
+from rasa.builder.constants import MAX_BACKUP_SIZE
+from rasa.builder.exceptions import ProjectGenerationError
 
 
 def _get_env_content() -> str:
@@ -138,3 +148,103 @@ def create_bot_project_archive(
 
     tar_buffer.seek(0)
     return tar_buffer.getvalue()
+
+
+def validate_s3_url(url: str) -> None:
+    """Validate that the URL is from an expected S3 domain for security.
+
+    Args:
+        url: The URL to validate
+
+    Raises:
+        ValueError: If the URL is not from an expected S3 domain
+    """
+    parsed = urlparse(url)
+    hostname = parsed.hostname
+
+    if not hostname:
+        raise ValueError("URL must have a valid hostname")
+
+    hostname = hostname.lower()
+    if not ("s3" in hostname and hostname.endswith(".amazonaws.com")):
+        raise ValueError(f"URL must be from an AWS S3 domain, got: {hostname}")
+
+
+async def download_backup_from_url(url: str) -> str:
+    """Download backup file from presigned URL to a temporary file.
+
+    Args:
+        url: Presigned URL to download from
+
+    Returns:
+        Path to the downloaded temporary file
+
+    Raises:
+        ProjectGenerationError: If download fails or file is too large
+    """
+    # Validate URL for security
+    validate_s3_url(url)
+
+    # Create temporary file path (using mktemp for path only, not creating the file)
+    temp_file_fd, temp_file_path = tempfile.mkstemp(suffix=".tar.gz")
+    os.close(temp_file_fd)  # Close the file descriptor immediately
+
+    try:
+        timeout = aiohttp.ClientTimeout(total=60)
+        async with aiohttp.ClientSession(timeout=timeout) as session:
+            async with session.get(url) as response:
+                if response.status != 200:
+                    raise ProjectGenerationError(
+                        f"Failed to download backup from presigned URL. "
+                        f"HTTP {response.status}: {response.reason}",
+                        attempts=1,
+                    )
+
+                # Check content length if available
+                content_length = response.headers.get("Content-Length")
+                if content_length and int(content_length) > MAX_BACKUP_SIZE:
+                    raise ProjectGenerationError(
+                        f"Backup file too large "
+                        f"({content_length} bytes > {MAX_BACKUP_SIZE} bytes). "
+                        f"Please provide a smaller backup file.",
+                        attempts=1,
+                    )
+
+                # Stream download to file using async file operations
+                downloaded_size = 0
+                async with aiofiles.open(temp_file_path, "wb") as f:
+                    async for chunk in response.content.iter_chunked(8192):
+                        downloaded_size += len(chunk)
+
+                        # Check size limit during download
+                        if downloaded_size > MAX_BACKUP_SIZE:
+                            raise ProjectGenerationError(
+                                f"Backup file too large "
+                                f"({downloaded_size} bytes > {MAX_BACKUP_SIZE} bytes).",
+                                attempts=1,
+                            )
+
+                        await f.write(chunk)
+
+                return temp_file_path
+
+    except ProjectGenerationError:
+        # Clean up temp file and re-raise ProjectGenerationError as-is
+        try:
+            Path(temp_file_path).unlink(missing_ok=True)
+        except Exception:
+            pass
+        raise
+    except asyncio.TimeoutError:
+        error_message = "Download timeout: Presigned URL may have expired."
+    except aiohttp.ClientError as exc:
+        error_message = f"Network error downloading backup: {exc}"
+    except Exception as exc:
+        error_message = f"Unexpected error downloading backup: {exc}"
+
+    # Clean up temp file and raise error
+    try:
+        Path(temp_file_path).unlink(missing_ok=True)
+    except Exception:
+        pass
+    raise ProjectGenerationError(error_message, attempts=1)

rasa_pro-3.15.0a3/rasa/builder/evaluator/constants.py

@@ -0,0 +1,15 @@
+"""Constants for the evaluator module."""
+
+from pathlib import Path
+
+# Base directory for the rasa package
+BASE_DIR = Path(__file__).parent.parent.parent
+
+# Response classification evaluation results directory
+RESPONSE_CLASSIFICATION_EVALUATION_RESULTS_DIR = (
+    BASE_DIR / "builder" / "evaluator" / "results"
+)
+# Default output filename
+DEFAULT_RESPONSE_CLASSIFICATION_EVALUATION_TEXT_OUTPUT_FILENAME = "run_results.txt"
+# Default YAML output filename
+RESPONSE_CLASSIFICATION_EVALUATION_YAML_OUTPUT_FILENAME = "run_results.yaml"

rasa_pro-3.15.0a3/rasa/builder/evaluator/copilot_executor.py

@@ -0,0 +1,89 @@
+"""Copilot execution utilities for evaluators.
+
+This module provides utilities for running copilot operations in evaluation contexts,
+independent of specific evaluation frameworks like Langfuse.
+"""
+
+from typing import List, Optional
+
+import structlog
+from pydantic import BaseModel
+
+from rasa.builder.config import COPILOT_HANDLER_ROLLING_BUFFER_SIZE
+from rasa.builder.copilot.models import (
+    CopilotContext,
+    CopilotGenerationContext,
+    GeneratedContent,
+    ReferenceSection,
+    ResponseCategory,
+)
+from rasa.builder.llm_service import llm_service
+
+structlogger = structlog.get_logger()
+
+
+class CopilotRunResult(BaseModel):
+    """Result from running the copilot with response handler."""
+
+    complete_response: Optional[str]
+    response_category: Optional[ResponseCategory]
+    reference_section: Optional[ReferenceSection]
+    generation_context: CopilotGenerationContext
+
+
+async def run_copilot_with_response_handler(
+    context: CopilotContext,
+) -> Optional[CopilotRunResult]:
+    """Run the copilot with response handler on the given context.
+
+    This function encapsulates the core copilot execution logic. It handles:
+    - Instantiating the copilot and response handler
+    - Generating a response and extracting the reference section from the given context
+    - Returning structured results
+
+    Args:
+        context: The copilot context to process.
+
+    Returns:
+        CopilotRunResult containing the complete response, category, and generation
+        context, or None if execution fails.
+
+    Raises:
+        Any exceptions from the copilot or response handler execution.
+    """
+    # Instantiate the copilot and response handler
+    copilot = llm_service.instantiate_copilot()
+    copilot_response_handler = llm_service.instantiate_handler(
+        COPILOT_HANDLER_ROLLING_BUFFER_SIZE
+    )
+
+    # Call the copilot to generate a response and handle it with the response
+    # handler
+    (original_stream, generation_context) = await copilot.generate_response(context)
+    intercepted_stream = copilot_response_handler.handle_response(original_stream)
+
+    # Exhaust the stream to get the complete response for evaluation
+    response_chunks: List[str] = []
+    response_category = None
+    async for chunk in intercepted_stream:
+        if not isinstance(chunk, GeneratedContent):
+            continue
+        response_chunks.append(chunk.content)
+        response_category = chunk.response_category
+
+    complete_response = "".join(response_chunks) if response_chunks else None
+
+    # Extract the reference section from the response handler
+    if generation_context.relevant_documents:
+        reference_section = copilot_response_handler.extract_references(
+            generation_context.relevant_documents
+        )
+    else:
+        reference_section = None
+
+    return CopilotRunResult(
+        complete_response=complete_response,
+        response_category=response_category,
+        reference_section=reference_section,
+        generation_context=generation_context,
+    )