rasa-pro 3.13.1a14__tar.gz → 3.13.1a16__tar.gz

{rasa_pro-3.13.1a14 → rasa_pro-3.13.1a16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: rasa-pro
-Version: 3.13.1a14
+Version: 3.13.1a16
 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
@@ -41,11 +41,11 @@ Requires-Dist: colorhash (>=2.0,<2.1.0)
 Requires-Dist: confluent-kafka (>=2.10.0,<3.0.0)
 Requires-Dist: cryptography (>=44.0.1)
 Requires-Dist: cvg-python-sdk (>=0.5.1,<0.6.0)
-Requires-Dist: dask (>=2024.7.0,<2024.8.0)
+Requires-Dist: dask (>=2024.8.0,<2024.9.0)
 Requires-Dist: demoji (>=1.1.0,<2.0.0)
 Requires-Dist: diskcache (>=5.6.3,<5.7.0)
 Requires-Dist: dnspython (==2.6.1)
-Requires-Dist: faiss-cpu (>=1.7.4,<2.0.0)
+Requires-Dist: faiss-cpu (>=1.11.0,<1.12.0)
 Requires-Dist: fbmessenger (>=6.0.0,<6.1.0)
 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"
@@ -65,7 +65,7 @@ Requires-Dist: langchain (>=0.2.17,<0.3.0)
 Requires-Dist: langchain-community (>=0.2.19,<0.3.0)
 Requires-Dist: langcodes (>=3.5.0,<4.0.0)
 Requires-Dist: litellm (>=1.69.0,<1.70.0)
-Requires-Dist: matplotlib (>=3.7,<3.8)
+Requires-Dist: matplotlib (>=3.9.4,<3.10.0)
 Requires-Dist: mattermostwrapper (>=2.2,<2.3)
 Requires-Dist: networkx (>=3.1,<3.2)
 Requires-Dist: numpy (>=1.26.4,<1.27.0)
@@ -99,7 +99,7 @@ Requires-Dist: pyyaml (>=6.0)
 Requires-Dist: qdrant-client (>=1.9.1,<1.10.0)
 Requires-Dist: questionary (>=1.10.0,<2.1.0)
 Requires-Dist: randomname (>=0.2.1,<0.3.0)
-Requires-Dist: rasa-sdk (==3.13.0rc1)
+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)
@@ -110,14 +110,15 @@ Requires-Dist: safetensors (>=0.4.5,<0.5.0)
 Requires-Dist: sanic (>=22.12,<22.13)
 Requires-Dist: sanic-cors (>=2.2.0,<2.3.0)
 Requires-Dist: sanic-jwt (>=1.8.0,<2.0.0)
+Requires-Dist: sanic-openapi (>=21.12.0,<22.0.0)
 Requires-Dist: sanic-routing (>=22.8.0,<23.0.0)
-Requires-Dist: scikit-learn (>=1.5.1,<1.6.0)
+Requires-Dist: scikit-learn (>=1.6.1,<1.7.0)
 Requires-Dist: scipy (>=1.13.1,<1.14.0)
 Requires-Dist: sentencepiece[sentencepiece] (>=0.1.99,<0.2.0) ; extra == "transformers" or extra == "full"
 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: sklearn-crfsuite (>=0.5.0,<0.6.0)
+Requires-Dist: skops (>=0.11.0,<0.12.0)
 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)
@@ -135,7 +136,7 @@ Requires-Dist: tensorflow-metal (==1.1.0) ; (sys_platform == "darwin" and platfo
 Requires-Dist: tensorflow-text (==2.14.0) ; sys_platform != "win32" and platform_machine != "arm64" and platform_machine != "aarch64"
 Requires-Dist: tensorflow_hub (>=0.13.0,<0.14.0)
 Requires-Dist: terminaltables (>=3.1.10,<3.2.0)
-Requires-Dist: tiktoken (>=0.7.0,<0.8.0)
+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"
 Requires-Dist: twilio (>=8.4,<8.5)

{rasa_pro-3.13.1a14 → rasa_pro-3.13.1a16}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "rasa-pro"
-version = "3.13.1a14"
+version = "3.13.1a16"
 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>",]
@@ -128,7 +128,7 @@ boto3 = "~1.35.15"
 google-cloud-storage = "~2.14.0"
 azure-storage-blob = ">=12.16.0, <12.17.0"
 requests = "~2.32.3"
-matplotlib = ">=3.7,<3.8"
+matplotlib = "~3.9.4"
 attrs = ">=23.1,<23.2"
 jsonpickle = ">=3.3.0,<3.4"
 redis = ">=4.6.0, <6.0"
@@ -163,7 +163,7 @@ python-socketio = ">=5.8,<6"
 python-engineio = "~4.12.2" # skip 5.0.0 upgrade
 pydot = "~1.4"
 SQLAlchemy = ">=2.0.41,<2.1.0"
-sklearn-crfsuite = "~0.3.6"
+sklearn-crfsuite = "~0.5.0"
 psycopg2-binary = ">=2.9.10,<2.10.0"
 python-dateutil = "~2.8.2"
 protobuf = "~4.25.8"
@@ -208,8 +208,8 @@ hvac = "^1.2.1"
 rich = "^13.4.2"
 cvg-python-sdk = "^0.5.1"
 python-keycloak = "^3.12.0"
-tiktoken = "~0.7.0"
-faiss-cpu = "^1.7.4"
+tiktoken = "~0.9.0"
+faiss-cpu = "~1.11.0"
 pymilvus = ">=2.4.1,<2.4.2" # pinned because 2.4.2 introduces milvus-lite dependency that fails on windows
 qdrant-client = "~1.9.1"
 psutil = "^5.9.5"
@@ -218,15 +218,16 @@ types-protobuf = "4.25.0.20240417"
 litellm = "~1.69.0"
 openpyxl = "^3.1.5"
 diskcache = "~5.6.3"
-skops = "~0.10.0"
+skops = "~0.11.0"
 safetensors = "~0.4.5"
-dask = "~2024.7.0" # later versions drop python 3.9 support
+dask = "~2024.8.0" # later versions drop python 3.9 support
 numpy = "~1.26.4"
 scipy = "~1.13.1"
-scikit-learn = "~1.5.1"
+scikit-learn = "~1.6.1"
 demoji = "^1.1.0"
 azure-identity = "~1.19.0"
 langcodes = "^3.5.0"
+sanic-openapi = "^21.12.0"
 
 [[tool.poetry.dependencies.gliner]]
 version = "~0.2.20"
@@ -301,7 +302,7 @@ timeout_func_only = true
 asyncio_mode = "auto"
 
 [tool.poetry.dependencies.rasa-sdk]
-version = "3.13.0rc1"
+version = "3.13.0"
 allow-prereleases = true
 
 [tool.poetry.dependencies.keras]
@@ -374,18 +375,6 @@ convention = "google"
 
 
 
-
-
-
-
-
-
-
-
-
-
-
-
 [tool.poetry.group.dev.dependencies]
 ruff = ">=0.3.5,<0.4.0"
 docker = "~7.1.0"

{rasa_pro-3.13.1a14 → rasa_pro-3.13.1a16}/rasa/builder/config.py

@@ -43,13 +43,18 @@ def get_default_config(assistant_id: str) -> Dict[str, Any]:
             )
         )
     )
+
+    if not isinstance(base_config, dict):
+        raise ValueError("Base config is not a dictionary")
+
     base_config["assistant_id"] = assistant_id
+
     return base_config
 
 
 def get_default_endpoints() -> Dict[str, Any]:
     """Get default endpoints configuration."""
-    return read_yaml_file(
+    endpoints_config = read_yaml_file(
         str(
             importlib_resources.files(PACKAGE_NAME).joinpath(
                 "cli/project_templates/default/endpoints.yml"
@@ -57,6 +62,11 @@ def get_default_endpoints() -> Dict[str, Any]:
         )
     )
 
+    if not isinstance(endpoints_config, dict):
+        raise ValueError("Endpoints config is not a dictionary")
+
+    return endpoints_config
+
 
 def get_default_credentials() -> Dict[str, Any]:
     """Get default credentials configuration."""

{rasa_pro-3.13.1a14 → rasa_pro-3.13.1a16}/rasa/builder/exceptions.py

@@ -29,6 +29,12 @@ class LLMGenerationError(PromptToBotError):
     pass
 
 
+class DocumentRetrievalError(PromptToBotError):
+    """Raised when document retrieval fails."""
+
+    pass
+
+
 class SchemaValidationError(PromptToBotError):
     """Raised when schema validation fails."""
 

rasa_pro-3.13.1a16/rasa/builder/inkeep-rag-response-schema.json

@@ -0,0 +1,64 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "properties": {
+    "content": {
+      "type": "array",
+      "items": {
+        "type": "object",
+        "properties": {
+          "type": {
+            "type": "string"
+          },
+          "source": {
+            "type": "object",
+            "properties": {
+              "content": {
+                "type": "array",
+                "items": {
+                  "type": "object",
+                  "properties": {
+                    "type": {
+                      "type": "string"
+                    },
+                    "media_type": {
+                      "type": "string"
+                    },
+                    "text": {
+                      "type": "string"
+                    },
+                    "data": {
+                      "type": "string"
+                    }
+                  }
+                }
+              },
+              "type": {
+                "type": "string"
+              },
+              "media_type": {
+                "type": "string"
+              },
+              "data": {
+                "type": "string"
+              }
+            }
+          },
+          "title": {
+            "type": "string"
+          },
+          "context": {
+            "type": "string"
+          },
+          "record_type": {
+            "type": "string"
+          },
+          "url": {
+            "type": "string"
+          }
+        }
+      }
+    }
+  },
+  "required": ["content"]
+} 

rasa_pro-3.13.1a16/rasa/builder/inkeep_document_retrieval.py

@@ -0,0 +1,212 @@
+import asyncio
+import json
+import os
+from contextlib import asynccontextmanager
+from typing import List, Optional
+
+import importlib_resources
+import openai
+import structlog
+from openai.types.chat import ChatCompletion
+
+from rasa.builder.exceptions import DocumentRetrievalError
+from rasa.builder.models import Document
+from rasa.constants import PACKAGE_NAME
+from rasa.shared.utils.io import read_json_file
+
+INKEEP_API_KEY_ENV_VAR = "INKEEP_API_KEY"
+INKEEP_RAG_RESPONSE_SCHEMA_PATH = str(
+    importlib_resources.files(PACKAGE_NAME).joinpath(
+        "builder/inkeep-rag-response-schema.json"
+    )
+)
+
+INKEEP_DOCUMENT_RETRIEVAL_MODEL = "inkeep-rag"
+INKEEP_BASE_URL = "https://api.inkeep.com/v1/"
+
+structlogger = structlog.get_logger()
+
+
+class InKeepDocumentRetrieval:
+    """Handles the document retrieval from InKeep AI."""
+
+    def __init__(
+        self,
+        api_key: Optional[str] = None,
+    ):
+        self._client: Optional[openai.AsyncOpenAI] = None
+        self._rag_schema = read_json_file(INKEEP_RAG_RESPONSE_SCHEMA_PATH)
+        self._api_key = api_key or os.getenv(INKEEP_API_KEY_ENV_VAR)
+
+    async def retrieve_documents(
+        self, query: str, temperature: float = 0.0, timeout: float = 30.0
+    ) -> List[Document]:
+        """Retrieve relevant documents using InKeep AI based on the given query.
+
+        Args:
+            query: The search query
+            temperature: Controls randomness in generation (0.0 for deterministic)
+            timeout: Timeout for the API call
+
+        Returns:
+            List of Document objects containing retrieved content
+
+        Raises:
+            LLMGenerationError: If the API call fails or returns invalid response
+        """
+        response = await self._call_inkeep_rag_api(
+            query=query,
+            temperature=temperature,
+            timeout=timeout,
+        )
+        documents = self._parse_documents_from_response(response)
+        return documents
+
+    async def _call_inkeep_rag_api(
+        self, query: str, temperature: float, timeout: float
+    ) -> ChatCompletion:
+        """Call InKeep AI RAG's API endpoint and return the response content.
+
+        Args:
+            query: The search query to send to InKeep
+            temperature: Controls randomness in generation (0.0 for deterministic)
+            timeout: Timeout for the API call
+
+        Returns:
+            The response content from InKeep AI. The response is made of the retrieved
+            documents.
+
+        Raises:
+            LLMGenerationError: If the API call fails or returns invalid response
+        """
+        request_params = {
+            "model": INKEEP_DOCUMENT_RETRIEVAL_MODEL,
+            "messages": [{"role": "user", "content": query}],
+            "temperature": temperature,
+            "timeout": timeout,
+            "response_format": {
+                "type": "json_schema",
+                "json_schema": self._rag_schema,
+            },
+        }
+        try:
+            async with self._get_client() as client:
+                response = await client.chat.completions.create(**request_params)
+
+                if not response.choices[0].message.content:
+                    structlogger.warning(
+                        "inkeep_document_retrieval.empty_response",
+                        event_info="InKeep AI returned an empty response. ",
+                        request_params=request_params,
+                        response_content=response.choices[0].message.content,
+                    )
+                    raise DocumentRetrievalError(
+                        "InKeep Document Retrieval: Empty response"
+                    )
+
+                return response
+
+        except openai.OpenAIError as e:
+            structlogger.error(
+                "inkeep_document_retrieval.api_error",
+                event_info="InKeep Document Retrieval: API error",
+                request_params=request_params,
+                error=e,
+            )
+            raise DocumentRetrievalError(f"InKeep Document Retrieval: API error: {e}")
+        except asyncio.TimeoutError as e:
+            structlogger.error(
+                "inkeep_document_retrieval.timeout_error",
+                event_info="InKeep Document Retrieval: Timeout error",
+                request_params=request_params,
+                error=e,
+            )
+            raise DocumentRetrievalError(f"InKeep AI request timed out: {e}")
+        except Exception as e:
+            structlogger.error(
+                "inkeep_document_retrieval.error",
+                event_info="InKeep Document Retrieval: Error",
+                request_params=request_params,
+                error=e,
+            )
+            raise DocumentRetrievalError(
+                f"InKeep Document Retrieval: Unexpected error: {e}"
+            )
+
+    @asynccontextmanager
+    async def _get_client(self):
+        """Get or create client that handles the API calls to InKeep AI."""
+        if self._client is None:
+            self._client = openai.AsyncOpenAI(
+                api_key=self._api_key,
+                base_url=INKEEP_BASE_URL,
+            )
+
+        try:
+            yield self._client
+        except Exception as e:
+            structlogger.error(
+                "inkeep_document_retrieval.client_error",
+                event_info="InKeep Document Retrieval: Client error",
+                error=str(e),
+            )
+            raise
+
+    def _parse_documents_from_response(
+        self, response: ChatCompletion
+    ) -> List[Document]:
+        """Parse the InKeep AI response into Document objects.
+
+        Args:
+            response: ChatCompletion response from InKeep AI's RAG model.
+
+        Returns:
+            List of Document objects
+        """
+        try:
+            content = response.choices[0].message.content
+            if not content:
+                return []
+
+            response_data = json.loads(content)
+            documents = []
+
+            for item in response_data.get("content", []):
+                try:
+                    document = Document.from_inkeep_rag_response(item)
+                    documents.append(document)
+                except Exception as e:
+                    structlogger.warning(
+                        "inkeep_document_retrieval.invalid_document_skipped",
+                        event_info=(
+                            "InKeep Document Retrieval: Invalid document structure "
+                            "skipped. Returning empty list for this item."
+                        ),
+                        error=str(e),
+                        item=item,
+                    )
+                    # Continue processing other items, skip this invalid one
+                    continue
+
+            return documents
+
+        except json.JSONDecodeError as e:
+            structlogger.warning(
+                "inkeep_document_retrieval.parse_response_failed",
+                event_info=(
+                    "InKeep Document Retrieval: Parse response failed. "
+                    "Returning empty list.",
+                ),
+                error=str(e),
+            )
+            return []
+        except Exception as e:
+            structlogger.error(
+                "inkeep_document_retrieval.parse_response_error",
+                event_info=(
+                    "InKeep Document Retrieval: Parse response error. "
+                    "Returning empty list.",
+                ),
+                error=str(e),
+            )
+            return []

{rasa_pro-3.13.1a14 → rasa_pro-3.13.1a16}/rasa/builder/llm_service.py

@@ -5,17 +5,19 @@ import importlib
 import json
 from contextlib import asynccontextmanager
 from copy import deepcopy
-from typing import Any, Dict, List, Optional
+from typing import Any, AsyncGenerator, Dict, List, Optional
 
 import importlib_resources
 import openai
 import structlog
 from jinja2 import Template
+from pydantic import ValidationError
 
 from rasa.builder import config
 from rasa.builder.exceptions import LLMGenerationError
+from rasa.builder.inkeep_document_retrieval import InKeepDocumentRetrieval
 from rasa.builder.llm_context import tracker_as_llm_context
-from rasa.builder.models import LLMBuilderContext
+from rasa.builder.models import Document, LLMBuilderContext, LLMHelperResponse
 from rasa.constants import PACKAGE_NAME
 from rasa.shared.constants import DOMAIN_SCHEMA_FILE, RESPONSES_SCHEMA_FILE
 from rasa.shared.core.flows.yaml_flows_io import FLOWS_SCHEMA_FILE
@@ -28,14 +30,14 @@ structlogger = structlog.get_logger()
 class LLMService:
     """Handles OpenAI LLM interactions with caching for efficiency."""
 
-    def __init__(self):
+    def __init__(self) -> None:
         self._client: Optional[openai.AsyncOpenAI] = None
         self._domain_schema: Optional[Dict[str, Any]] = None
         self._flows_schema: Optional[Dict[str, Any]] = None
         self._helper_schema: Optional[Dict[str, Any]] = None
 
     @asynccontextmanager
-    async def _get_client(self):
+    async def _get_client(self) -> AsyncGenerator[openai.AsyncOpenAI, None]:
         """Get or create OpenAI client with proper resource management."""
         if self._client is None:
             self._client = openai.AsyncOpenAI(timeout=config.OPENAI_TIMEOUT)
@@ -46,7 +48,7 @@ class LLMService:
             structlogger.error("llm.client_error", error=str(e))
             raise
 
-    def _prepare_schemas(self):
+    def _prepare_schemas(self) -> None:
         """Prepare and cache schemas for LLM generation."""
         if self._domain_schema is None:
             self._domain_schema = _prepare_domain_schema()
@@ -135,7 +137,7 @@ class LLMService:
 
     async def generate_helper_response(
         self, messages: List[Dict[str, Any]]
-    ) -> Dict[str, Any]:
+    ) -> LLMHelperResponse:
         """Generate helper response using OpenAI."""
         self._prepare_schemas()
 
@@ -158,9 +160,11 @@ class LLMService:
                     raise LLMGenerationError("Empty response from LLM helper")
 
                 try:
-                    return json.loads(content)
+                    return LLMHelperResponse.model_validate_json(json.loads(content))
                 except json.JSONDecodeError as e:
                     raise LLMGenerationError(f"Invalid JSON from LLM helper: {e}")
+                except ValidationError as e:
+                    raise LLMGenerationError(f"Invalid JSON from LLM helper: {e}")
 
         except openai.OpenAIError as e:
             raise LLMGenerationError(f"OpenAI API error in helper: {e}")
@@ -169,30 +173,14 @@ class LLMService:
 
     async def search_documentation(
         self, query: str, max_results: Optional[int] = None
-    ) -> List[Dict[str, Any]]:
+    ) -> List[Document]:
         """Search documentation using OpenAI vector store."""
-        if max_results is None:
-            max_results = config.OPENAI_MAX_VECTOR_RESULTS
-
-        try:
-            async with self._get_client() as client:
-                results = await client.vector_stores.search(
-                    vector_store_id=config.OPENAI_VECTOR_STORE_ID,
-                    query=query,
-                    max_num_results=max_results,
-                    rewrite_query=True,
-                )
-
-                return results.data
-
-        except openai.OpenAIError as e:
-            structlogger.warning(
-                "llm.documentation_search_failed", error=str(e), query=query
-            )
-            return []
+        inkeep_document_retrieval = InKeepDocumentRetrieval()
+        documents = await inkeep_document_retrieval.retrieve_documents(query)
+        return documents
 
     @staticmethod
-    def _format_chat_dump(messages) -> str:
+    def _format_chat_dump(messages: List[Dict[str, Any]]) -> str:
         """Format chat messages for documentation search."""
         result = ""
         for message in messages:
@@ -213,16 +201,15 @@ class LLMService:
         return result
 
     @staticmethod
-    def _format_documentation_results(results) -> str:
+    def _format_documentation_results(results: List[Document]) -> str:
         """Format documentation search results."""
         if not results:
             return "<sources>No relevant documentation found.</sources>"
 
         formatted_results = ""
         for result in results:
-            formatted_result = f"<result url='{result.attributes.get('url', '')}'>"
-            for part in result.content:
-                formatted_result += f"<content>{part.text}</content>"
+            formatted_result = f"<result url='{result.url}'>"
+            formatted_result += f"<content>{result.content}</content>"
             formatted_results += formatted_result + "</result>"
 
         return f"<sources>{formatted_results}</sources>"
@@ -233,6 +220,9 @@ def _prepare_domain_schema() -> Dict[str, Any]:
     """Prepare domain schema by removing unnecessary parts."""
     domain_schema = deepcopy(read_schema_file(DOMAIN_SCHEMA_FILE, PACKAGE_NAME, False))
 
+    if not isinstance(domain_schema, dict):
+        raise ValueError("Domain schema is not a dictionary")
+
     # Remove parts not needed for CALM bots
     unnecessary_keys = ["intents", "entities", "forms", "config", "session_config"]