langflow-base-nightly 0.6.5.dev4__py3-none-any.whl → 0.6.5.dev6__py3-none-any.whl

langflow/initial_setup/starter_projects/Hybrid Search RAG.json

@@ -734,7 +734,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -2695,7 +2695,7 @@
             "legacy": false,
             "lf_version": "1.6.0",
             "metadata": {
-              "code_hash": "a5b1b04d9fcc",
+              "code_hash": "9ad8d8d8fc16",
               "dependencies": {
                 "dependencies": [
                   {
@@ -2766,7 +2766,7 @@
                 "show": true,
                 "title_case": false,
                 "type": "code",
-                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(list[output_model_], Field(description=f\"A list of {schema_name}.\")),  # type: ignore[valid-type]\n        )\n\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[output_model])\n        except NotImplementedError as exc:\n            msg = f\"{self.llm.__class__.__name__} does not support structured output.\"\n            raise TypeError(msg) from exc\n\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        result = get_chat_result(\n            runnable=llm_with_structured_output,\n            system_message=self.system_prompt,\n            input_value=self.input_value,\n            config=config_dict,\n        )\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response.model_dump() if isinstance(first_response, BaseModel) else first_response\n\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n"
+                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.log.logger import logger\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(\n                list[output_model_],\n                Field(\n                    description=f\"A list of {schema_name}.\",  # type: ignore[valid-type]\n                    min_length=1,  # help ensure non-empty output\n                ),\n            ),\n        )\n        # Tracing config\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        # Generate structured output using Trustcall first, then fallback to Langchain if it fails\n        result = self._extract_output_with_trustcall(output_model, config_dict)\n        if result is None:\n            result = self._extract_output_with_langchain(output_model, config_dict)\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response\n        if isinstance(first_response, BaseModel):\n            structured_data = first_response.model_dump()\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n\n    def _extract_output_with_trustcall(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[schema], tool_choice=schema.__name__)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n        except Exception as e:  # noqa: BLE001\n            logger.warning(\n                f\"Trustcall extraction failed, falling back to Langchain: {e} \"\n                \"(Note: This may not be an error—some models or configurations do not support tool calling. \"\n                \"Falling back is normal in such cases.)\"\n            )\n            return None\n        return result or None  # langchain fallback is used if error occurs or the result is empty\n\n    def _extract_output_with_langchain(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = self.llm.with_structured_output(schema)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n            if isinstance(result, BaseModel):\n                result = result.model_dump()\n                result = result.get(\"objects\", result)\n        except Exception as fallback_error:\n            msg = (\n                f\"Model does not support tool calling (trustcall failed) \"\n                f\"and fallback with_structured_output also failed: {fallback_error}\"\n            )\n            raise ValueError(msg) from fallback_error\n\n        return result or None\n"
               },
               "input_value": {
                 "_input_type": "MultilineInput",

langflow/initial_setup/starter_projects/Image Sentiment Analysis.json

@@ -517,7 +517,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -1775,7 +1775,7 @@
             "legacy": false,
             "lf_version": "1.6.0",
             "metadata": {
-              "code_hash": "a5b1b04d9fcc",
+              "code_hash": "9ad8d8d8fc16",
               "dependencies": {
                 "dependencies": [
                   {
@@ -1846,7 +1846,7 @@
                 "show": true,
                 "title_case": false,
                 "type": "code",
-                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(list[output_model_], Field(description=f\"A list of {schema_name}.\")),  # type: ignore[valid-type]\n        )\n\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[output_model])\n        except NotImplementedError as exc:\n            msg = f\"{self.llm.__class__.__name__} does not support structured output.\"\n            raise TypeError(msg) from exc\n\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        result = get_chat_result(\n            runnable=llm_with_structured_output,\n            system_message=self.system_prompt,\n            input_value=self.input_value,\n            config=config_dict,\n        )\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response.model_dump() if isinstance(first_response, BaseModel) else first_response\n\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n"
+                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.log.logger import logger\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(\n                list[output_model_],\n                Field(\n                    description=f\"A list of {schema_name}.\",  # type: ignore[valid-type]\n                    min_length=1,  # help ensure non-empty output\n                ),\n            ),\n        )\n        # Tracing config\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        # Generate structured output using Trustcall first, then fallback to Langchain if it fails\n        result = self._extract_output_with_trustcall(output_model, config_dict)\n        if result is None:\n            result = self._extract_output_with_langchain(output_model, config_dict)\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response\n        if isinstance(first_response, BaseModel):\n            structured_data = first_response.model_dump()\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n\n    def _extract_output_with_trustcall(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[schema], tool_choice=schema.__name__)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n        except Exception as e:  # noqa: BLE001\n            logger.warning(\n                f\"Trustcall extraction failed, falling back to Langchain: {e} \"\n                \"(Note: This may not be an error—some models or configurations do not support tool calling. \"\n                \"Falling back is normal in such cases.)\"\n            )\n            return None\n        return result or None  # langchain fallback is used if error occurs or the result is empty\n\n    def _extract_output_with_langchain(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = self.llm.with_structured_output(schema)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n            if isinstance(result, BaseModel):\n                result = result.model_dump()\n                result = result.get(\"objects\", result)\n        except Exception as fallback_error:\n            msg = (\n                f\"Model does not support tool calling (trustcall failed) \"\n                f\"and fallback with_structured_output also failed: {fallback_error}\"\n            )\n            raise ValueError(msg) from fallback_error\n\n        return result or None\n"
               },
               "input_value": {
                 "_input_type": "MultilineInput",

langflow/initial_setup/starter_projects/Instagram Copywriter.json

@@ -1051,7 +1051,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Invoice Summarizer.json

@@ -314,7 +314,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Knowledge Retrieval.json

@@ -241,7 +241,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Market Research.json

@@ -485,7 +485,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -2387,7 +2387,7 @@
             "legacy": false,
             "lf_version": "1.6.0",
             "metadata": {
-              "code_hash": "a5b1b04d9fcc",
+              "code_hash": "9ad8d8d8fc16",
               "dependencies": {
                 "dependencies": [
                   {
@@ -2458,7 +2458,7 @@
                 "show": true,
                 "title_case": false,
                 "type": "code",
-                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(list[output_model_], Field(description=f\"A list of {schema_name}.\")),  # type: ignore[valid-type]\n        )\n\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[output_model])\n        except NotImplementedError as exc:\n            msg = f\"{self.llm.__class__.__name__} does not support structured output.\"\n            raise TypeError(msg) from exc\n\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        result = get_chat_result(\n            runnable=llm_with_structured_output,\n            system_message=self.system_prompt,\n            input_value=self.input_value,\n            config=config_dict,\n        )\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response.model_dump() if isinstance(first_response, BaseModel) else first_response\n\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n"
+                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.log.logger import logger\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(\n                list[output_model_],\n                Field(\n                    description=f\"A list of {schema_name}.\",  # type: ignore[valid-type]\n                    min_length=1,  # help ensure non-empty output\n                ),\n            ),\n        )\n        # Tracing config\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        # Generate structured output using Trustcall first, then fallback to Langchain if it fails\n        result = self._extract_output_with_trustcall(output_model, config_dict)\n        if result is None:\n            result = self._extract_output_with_langchain(output_model, config_dict)\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response\n        if isinstance(first_response, BaseModel):\n            structured_data = first_response.model_dump()\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n\n    def _extract_output_with_trustcall(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[schema], tool_choice=schema.__name__)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n        except Exception as e:  # noqa: BLE001\n            logger.warning(\n                f\"Trustcall extraction failed, falling back to Langchain: {e} \"\n                \"(Note: This may not be an error—some models or configurations do not support tool calling. \"\n                \"Falling back is normal in such cases.)\"\n            )\n            return None\n        return result or None  # langchain fallback is used if error occurs or the result is empty\n\n    def _extract_output_with_langchain(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = self.llm.with_structured_output(schema)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n            if isinstance(result, BaseModel):\n                result = result.model_dump()\n                result = result.get(\"objects\", result)\n        except Exception as fallback_error:\n            msg = (\n                f\"Model does not support tool calling (trustcall failed) \"\n                f\"and fallback with_structured_output also failed: {fallback_error}\"\n            )\n            raise ValueError(msg) from fallback_error\n\n        return result or None\n"
               },
               "input_value": {
                 "_input_type": "MultilineInput",

langflow/initial_setup/starter_projects/Meeting Summary.json

@@ -648,7 +648,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -927,7 +927,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -1206,7 +1206,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Memory Chatbot.json

@@ -432,7 +432,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/News Aggregator.json

@@ -889,7 +889,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -1847,7 +1847,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -1859,7 +1859,7 @@
                   },
                   {
                     "name": "boto3",
-                    "version": "1.40.52"
+                    "version": "1.40.56"
                   },
                   {
                     "name": "google",

langflow/initial_setup/starter_projects/Nvidia Remix.json

@@ -520,7 +520,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -1804,7 +1804,7 @@
                   },
                   {
                     "name": "langchain_ollama",
-                    "version": "0.2.1"
+                    "version": "0.3.10"
                   },
                   {
                     "name": "langchain_community",
@@ -2483,7 +2483,7 @@
             "legacy": false,
             "lf_version": "1.4.2",
             "metadata": {
-              "code_hash": "23cc4057ba86",
+              "code_hash": "81b1338be859",
               "dependencies": {
                 "dependencies": [
                   {
@@ -2543,7 +2543,7 @@
                 "show": true,
                 "title_case": false,
                 "type": "code",
-                "value": "from __future__ import annotations\n\nimport asyncio\nimport uuid\nfrom typing import Any\n\nfrom langchain_core.tools import StructuredTool  # noqa: TC002\n\nfrom lfx.base.agents.utils import maybe_unflatten_dict, safe_cache_get, safe_cache_set\nfrom lfx.base.mcp.util import (\n    MCPStdioClient,\n    MCPStreamableHttpClient,\n    create_input_schema_from_json_schema,\n    update_tools,\n)\nfrom lfx.custom.custom_component.component_with_cache import ComponentWithCache\nfrom lfx.inputs.inputs import InputTypes  # noqa: TC001\nfrom lfx.io import BoolInput, DropdownInput, McpInput, MessageTextInput, Output\nfrom lfx.io.schema import flatten_schema, schema_to_langflow_inputs\nfrom lfx.log.logger import logger\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.message import Message\nfrom lfx.services.deps import get_settings_service, get_storage_service, session_scope\n\n\nclass MCPToolsComponent(ComponentWithCache):\n    schema_inputs: list = []\n    tools: list[StructuredTool] = []\n    _not_load_actions: bool = False\n    _tool_cache: dict = {}\n    _last_selected_server: str | None = None  # Cache for the last selected server\n\n    def __init__(self, **data) -> None:\n        super().__init__(**data)\n        # Initialize cache keys to avoid CacheMiss when accessing them\n        self._ensure_cache_structure()\n\n        # Initialize clients with access to the component cache\n        self.stdio_client: MCPStdioClient = MCPStdioClient(component_cache=self._shared_component_cache)\n        self.streamable_http_client: MCPStreamableHttpClient = MCPStreamableHttpClient(\n            component_cache=self._shared_component_cache\n        )\n\n    def _ensure_cache_structure(self):\n        \"\"\"Ensure the cache has the required structure.\"\"\"\n        # Check if servers key exists and is not CacheMiss\n        servers_value = safe_cache_get(self._shared_component_cache, \"servers\")\n        if servers_value is None:\n            safe_cache_set(self._shared_component_cache, \"servers\", {})\n\n        # Check if last_selected_server key exists and is not CacheMiss\n        last_server_value = safe_cache_get(self._shared_component_cache, \"last_selected_server\")\n        if last_server_value is None:\n            safe_cache_set(self._shared_component_cache, \"last_selected_server\", \"\")\n\n    default_keys: list[str] = [\n        \"code\",\n        \"_type\",\n        \"tool_mode\",\n        \"tool_placeholder\",\n        \"mcp_server\",\n        \"tool\",\n        \"use_cache\",\n        \"verify_ssl\",\n    ]\n\n    display_name = \"MCP Tools\"\n    description = \"Connect to an MCP server to use its tools.\"\n    documentation: str = \"https://docs.langflow.org/mcp-client\"\n    icon = \"Mcp\"\n    name = \"MCPTools\"\n\n    inputs = [\n        McpInput(\n            name=\"mcp_server\",\n            display_name=\"MCP Server\",\n            info=\"Select the MCP Server that will be used by this component\",\n            real_time_refresh=True,\n        ),\n        BoolInput(\n            name=\"use_cache\",\n            display_name=\"Use Cached Server\",\n            info=(\n                \"Enable caching of MCP Server and tools to improve performance. \"\n                \"Disable to always fetch fresh tools and server updates.\"\n            ),\n            value=False,\n            advanced=True,\n        ),\n        BoolInput(\n            name=\"verify_ssl\",\n            display_name=\"Verify SSL Certificate\",\n            info=(\n                \"Enable SSL certificate verification for HTTPS connections. \"\n                \"Disable only for development/testing with self-signed certificates.\"\n            ),\n            value=True,\n            advanced=True,\n        ),\n        DropdownInput(\n            name=\"tool\",\n            display_name=\"Tool\",\n            options=[],\n            value=\"\",\n            info=\"Select the tool to execute\",\n            show=False,\n            required=True,\n            real_time_refresh=True,\n        ),\n        MessageTextInput(\n            name=\"tool_placeholder\",\n            display_name=\"Tool Placeholder\",\n            info=\"Placeholder for the tool\",\n            value=\"\",\n            show=False,\n            tool_mode=False,\n        ),\n    ]\n\n    outputs = [\n        Output(display_name=\"Response\", name=\"response\", method=\"build_output\"),\n    ]\n\n    async def _validate_schema_inputs(self, tool_obj) -> list[InputTypes]:\n        \"\"\"Validate and process schema inputs for a tool.\"\"\"\n        try:\n            if not tool_obj or not hasattr(tool_obj, \"args_schema\"):\n                msg = \"Invalid tool object or missing input schema\"\n                raise ValueError(msg)\n\n            flat_schema = flatten_schema(tool_obj.args_schema.schema())\n            input_schema = create_input_schema_from_json_schema(flat_schema)\n            if not input_schema:\n                msg = f\"Empty input schema for tool '{tool_obj.name}'\"\n                raise ValueError(msg)\n\n            schema_inputs = schema_to_langflow_inputs(input_schema)\n            if not schema_inputs:\n                msg = f\"No input parameters defined for tool '{tool_obj.name}'\"\n                await logger.awarning(msg)\n                return []\n\n        except Exception as e:\n            msg = f\"Error validating schema inputs: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return schema_inputs\n\n    async def update_tool_list(self, mcp_server_value=None):\n        # Accepts mcp_server_value as dict {name, config} or uses self.mcp_server\n        mcp_server = mcp_server_value if mcp_server_value is not None else getattr(self, \"mcp_server\", None)\n        server_name = None\n        server_config_from_value = None\n        if isinstance(mcp_server, dict):\n            server_name = mcp_server.get(\"name\")\n            server_config_from_value = mcp_server.get(\"config\")\n        else:\n            server_name = mcp_server\n        if not server_name:\n            self.tools = []\n            return [], {\"name\": server_name, \"config\": server_config_from_value}\n\n        # Check if caching is enabled, default to False\n        use_cache = getattr(self, \"use_cache\", False)\n\n        # Use shared cache if available and caching is enabled\n        cached = None\n        if use_cache:\n            servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n            cached = servers_cache.get(server_name) if isinstance(servers_cache, dict) else None\n\n        if cached is not None:\n            try:\n                self.tools = cached[\"tools\"]\n                self.tool_names = cached[\"tool_names\"]\n                self._tool_cache = cached[\"tool_cache\"]\n                server_config_from_value = cached[\"config\"]\n            except (TypeError, KeyError, AttributeError) as e:\n                # Handle corrupted cache data by clearing it and continuing to fetch fresh tools\n                msg = f\"Unable to use cached data for MCP Server{server_name}: {e}\"\n                await logger.awarning(msg)\n                # Clear the corrupted cache entry\n                current_servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                if isinstance(current_servers_cache, dict) and server_name in current_servers_cache:\n                    current_servers_cache.pop(server_name)\n                    safe_cache_set(self._shared_component_cache, \"servers\", current_servers_cache)\n            else:\n                return self.tools, {\"name\": server_name, \"config\": server_config_from_value}\n\n        try:\n            try:\n                from langflow.api.v2.mcp import get_server\n                from langflow.services.database.models.user.crud import get_user_by_id\n            except ImportError as e:\n                msg = (\n                    \"Langflow MCP server functionality is not available. \"\n                    \"This feature requires the full Langflow installation.\"\n                )\n                raise ImportError(msg) from e\n            async with session_scope() as db:\n                if not self.user_id:\n                    msg = \"User ID is required for fetching MCP tools.\"\n                    raise ValueError(msg)\n                current_user = await get_user_by_id(db, self.user_id)\n\n                # Try to get server config from DB/API\n                server_config = await get_server(\n                    server_name,\n                    current_user,\n                    db,\n                    storage_service=get_storage_service(),\n                    settings_service=get_settings_service(),\n                )\n\n            # If get_server returns empty but we have a config, use it\n            if not server_config and server_config_from_value:\n                server_config = server_config_from_value\n\n            if not server_config:\n                self.tools = []\n                return [], {\"name\": server_name, \"config\": server_config}\n\n            # Add verify_ssl option to server config if not present\n            if \"verify_ssl\" not in server_config:\n                verify_ssl = getattr(self, \"verify_ssl\", True)\n                server_config[\"verify_ssl\"] = verify_ssl\n\n            _, tool_list, tool_cache = await update_tools(\n                server_name=server_name,\n                server_config=server_config,\n                mcp_stdio_client=self.stdio_client,\n                mcp_streamable_http_client=self.streamable_http_client,\n            )\n\n            self.tool_names = [tool.name for tool in tool_list if hasattr(tool, \"name\")]\n            self._tool_cache = tool_cache\n            self.tools = tool_list\n\n            # Cache the result only if caching is enabled\n            if use_cache:\n                cache_data = {\n                    \"tools\": tool_list,\n                    \"tool_names\": self.tool_names,\n                    \"tool_cache\": tool_cache,\n                    \"config\": server_config,\n                }\n\n                # Safely update the servers cache\n                current_servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                if isinstance(current_servers_cache, dict):\n                    current_servers_cache[server_name] = cache_data\n                    safe_cache_set(self._shared_component_cache, \"servers\", current_servers_cache)\n\n        except (TimeoutError, asyncio.TimeoutError) as e:\n            msg = f\"Timeout updating tool list: {e!s}\"\n            await logger.aexception(msg)\n            raise TimeoutError(msg) from e\n        except Exception as e:\n            msg = f\"Error updating tool list: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return tool_list, {\"name\": server_name, \"config\": server_config}\n\n    async def update_build_config(self, build_config: dict, field_value: str, field_name: str | None = None) -> dict:\n        \"\"\"Toggle the visibility of connection-specific fields based on the selected mode.\"\"\"\n        try:\n            if field_name == \"tool\":\n                try:\n                    if len(self.tools) == 0:\n                        try:\n                            self.tools, build_config[\"mcp_server\"][\"value\"] = await self.update_tool_list()\n                            build_config[\"tool\"][\"options\"] = [tool.name for tool in self.tools]\n                            build_config[\"tool\"][\"placeholder\"] = \"Select a tool\"\n                        except (TimeoutError, asyncio.TimeoutError) as e:\n                            msg = f\"Timeout updating tool list: {e!s}\"\n                            await logger.aexception(msg)\n                            if not build_config[\"tools_metadata\"][\"show\"]:\n                                build_config[\"tool\"][\"show\"] = True\n                                build_config[\"tool\"][\"options\"] = []\n                                build_config[\"tool\"][\"value\"] = \"\"\n                                build_config[\"tool\"][\"placeholder\"] = \"Timeout on MCP server\"\n                            else:\n                                build_config[\"tool\"][\"show\"] = False\n                        except ValueError:\n                            if not build_config[\"tools_metadata\"][\"show\"]:\n                                build_config[\"tool\"][\"show\"] = True\n                                build_config[\"tool\"][\"options\"] = []\n                                build_config[\"tool\"][\"value\"] = \"\"\n                                build_config[\"tool\"][\"placeholder\"] = \"Error on MCP Server\"\n                            else:\n                                build_config[\"tool\"][\"show\"] = False\n\n                    if field_value == \"\":\n                        return build_config\n                    tool_obj = None\n                    for tool in self.tools:\n                        if tool.name == field_value:\n                            tool_obj = tool\n                            break\n                    if tool_obj is None:\n                        msg = f\"Tool {field_value} not found in available tools: {self.tools}\"\n                        await logger.awarning(msg)\n                        return build_config\n                    await self._update_tool_config(build_config, field_value)\n                except Exception as e:\n                    build_config[\"tool\"][\"options\"] = []\n                    msg = f\"Failed to update tools: {e!s}\"\n                    raise ValueError(msg) from e\n                else:\n                    return build_config\n            elif field_name == \"mcp_server\":\n                if not field_value:\n                    build_config[\"tool\"][\"show\"] = False\n                    build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"value\"] = \"\"\n                    build_config[\"tool\"][\"placeholder\"] = \"\"\n                    build_config[\"tool_placeholder\"][\"tool_mode\"] = False\n                    self.remove_non_default_keys(build_config)\n                    return build_config\n\n                build_config[\"tool_placeholder\"][\"tool_mode\"] = True\n\n                current_server_name = field_value.get(\"name\") if isinstance(field_value, dict) else field_value\n                _last_selected_server = safe_cache_get(self._shared_component_cache, \"last_selected_server\", \"\")\n\n                # To avoid unnecessary updates, only proceed if the server has actually changed\n                if (_last_selected_server in (current_server_name, \"\")) and build_config[\"tool\"][\"show\"]:\n                    if current_server_name:\n                        servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                        if isinstance(servers_cache, dict):\n                            cached = servers_cache.get(current_server_name)\n                            if cached is not None and cached.get(\"tool_names\"):\n                                cached_tools = cached[\"tool_names\"]\n                                current_tools = build_config[\"tool\"][\"options\"]\n                                if current_tools == cached_tools:\n                                    return build_config\n                    else:\n                        return build_config\n\n                # Determine if \"Tool Mode\" is active by checking if the tool dropdown is hidden.\n                is_in_tool_mode = build_config[\"tools_metadata\"][\"show\"]\n                safe_cache_set(self._shared_component_cache, \"last_selected_server\", current_server_name)\n\n                # Check if tools are already cached for this server before clearing\n                cached_tools = None\n                if current_server_name:\n                    use_cache = getattr(self, \"use_cache\", True)\n                    if use_cache:\n                        servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                        if isinstance(servers_cache, dict):\n                            cached = servers_cache.get(current_server_name)\n                            if cached is not None:\n                                try:\n                                    cached_tools = cached[\"tools\"]\n                                    self.tools = cached_tools\n                                    self.tool_names = cached[\"tool_names\"]\n                                    self._tool_cache = cached[\"tool_cache\"]\n                                except (TypeError, KeyError, AttributeError) as e:\n                                    # Handle corrupted cache data by ignoring it\n                                    msg = f\"Unable to use cached data for MCP Server,{current_server_name}: {e}\"\n                                    await logger.awarning(msg)\n                                    cached_tools = None\n\n                # Only clear tools if we don't have cached tools for the current server\n                if not cached_tools:\n                    self.tools = []  # Clear previous tools only if no cache\n\n                self.remove_non_default_keys(build_config)  # Clear previous tool inputs\n\n                # Only show the tool dropdown if not in tool_mode\n                if not is_in_tool_mode:\n                    build_config[\"tool\"][\"show\"] = True\n                    if cached_tools:\n                        # Use cached tools to populate options immediately\n                        build_config[\"tool\"][\"options\"] = [tool.name for tool in cached_tools]\n                        build_config[\"tool\"][\"placeholder\"] = \"Select a tool\"\n                    else:\n                        # Show loading state only when we need to fetch tools\n                        build_config[\"tool\"][\"placeholder\"] = \"Loading tools...\"\n                        build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"value\"] = uuid.uuid4()\n                else:\n                    # Keep the tool dropdown hidden if in tool_mode\n                    self._not_load_actions = True\n                    build_config[\"tool\"][\"show\"] = False\n\n            elif field_name == \"tool_mode\":\n                build_config[\"tool\"][\"placeholder\"] = \"\"\n                build_config[\"tool\"][\"show\"] = not bool(field_value) and bool(build_config[\"mcp_server\"])\n                self.remove_non_default_keys(build_config)\n                self.tool = build_config[\"tool\"][\"value\"]\n                if field_value:\n                    self._not_load_actions = True\n                else:\n                    build_config[\"tool\"][\"value\"] = uuid.uuid4()\n                    build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"show\"] = True\n                    build_config[\"tool\"][\"placeholder\"] = \"Loading tools...\"\n            elif field_name == \"tools_metadata\":\n                self._not_load_actions = False\n\n        except Exception as e:\n            msg = f\"Error in update_build_config: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return build_config\n\n    def get_inputs_for_all_tools(self, tools: list) -> dict:\n        \"\"\"Get input schemas for all tools.\"\"\"\n        inputs = {}\n        for tool in tools:\n            if not tool or not hasattr(tool, \"name\"):\n                continue\n            try:\n                flat_schema = flatten_schema(tool.args_schema.schema())\n                input_schema = create_input_schema_from_json_schema(flat_schema)\n                langflow_inputs = schema_to_langflow_inputs(input_schema)\n                inputs[tool.name] = langflow_inputs\n            except (AttributeError, ValueError, TypeError, KeyError) as e:\n                msg = f\"Error getting inputs for tool {getattr(tool, 'name', 'unknown')}: {e!s}\"\n                logger.exception(msg)\n                continue\n        return inputs\n\n    def remove_input_schema_from_build_config(\n        self, build_config: dict, tool_name: str, input_schema: dict[list[InputTypes], Any]\n    ):\n        \"\"\"Remove the input schema for the tool from the build config.\"\"\"\n        # Keep only schemas that don't belong to the current tool\n        input_schema = {k: v for k, v in input_schema.items() if k != tool_name}\n        # Remove all inputs from other tools\n        for value in input_schema.values():\n            for _input in value:\n                if _input.name in build_config:\n                    build_config.pop(_input.name)\n\n    def remove_non_default_keys(self, build_config: dict) -> None:\n        \"\"\"Remove non-default keys from the build config.\"\"\"\n        for key in list(build_config.keys()):\n            if key not in self.default_keys:\n                build_config.pop(key)\n\n    async def _update_tool_config(self, build_config: dict, tool_name: str) -> None:\n        \"\"\"Update tool configuration with proper error handling.\"\"\"\n        if not self.tools:\n            self.tools, build_config[\"mcp_server\"][\"value\"] = await self.update_tool_list()\n\n        if not tool_name:\n            return\n\n        tool_obj = next((tool for tool in self.tools if tool.name == tool_name), None)\n        if not tool_obj:\n            msg = f\"Tool {tool_name} not found in available tools: {self.tools}\"\n            self.remove_non_default_keys(build_config)\n            build_config[\"tool\"][\"value\"] = \"\"\n            await logger.awarning(msg)\n            return\n\n        try:\n            # Store current values before removing inputs\n            current_values = {}\n            for key, value in build_config.items():\n                if key not in self.default_keys and isinstance(value, dict) and \"value\" in value:\n                    current_values[key] = value[\"value\"]\n\n            # Get all tool inputs and remove old ones\n            input_schema_for_all_tools = self.get_inputs_for_all_tools(self.tools)\n            self.remove_input_schema_from_build_config(build_config, tool_name, input_schema_for_all_tools)\n\n            # Get and validate new inputs\n            self.schema_inputs = await self._validate_schema_inputs(tool_obj)\n            if not self.schema_inputs:\n                msg = f\"No input parameters to configure for tool '{tool_name}'\"\n                await logger.ainfo(msg)\n                return\n\n            # Add new inputs to build config\n            for schema_input in self.schema_inputs:\n                if not schema_input or not hasattr(schema_input, \"name\"):\n                    msg = \"Invalid schema input detected, skipping\"\n                    await logger.awarning(msg)\n                    continue\n\n                try:\n                    name = schema_input.name\n                    input_dict = schema_input.to_dict()\n                    input_dict.setdefault(\"value\", None)\n                    input_dict.setdefault(\"required\", True)\n\n                    build_config[name] = input_dict\n\n                    # Preserve existing value if the parameter name exists in current_values\n                    if name in current_values:\n                        build_config[name][\"value\"] = current_values[name]\n\n                except (AttributeError, KeyError, TypeError) as e:\n                    msg = f\"Error processing schema input {schema_input}: {e!s}\"\n                    await logger.aexception(msg)\n                    continue\n        except ValueError as e:\n            msg = f\"Schema validation error for tool {tool_name}: {e!s}\"\n            await logger.aexception(msg)\n            self.schema_inputs = []\n            return\n        except (AttributeError, KeyError, TypeError) as e:\n            msg = f\"Error updating tool config: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n\n    async def build_output(self) -> DataFrame:\n        \"\"\"Build output with improved error handling and validation.\"\"\"\n        try:\n            self.tools, _ = await self.update_tool_list()\n            if self.tool != \"\":\n                # Set session context for persistent MCP sessions using Langflow session ID\n                session_context = self._get_session_context()\n                if session_context:\n                    self.stdio_client.set_session_context(session_context)\n                    self.streamable_http_client.set_session_context(session_context)\n\n                exec_tool = self._tool_cache[self.tool]\n                tool_args = self.get_inputs_for_all_tools(self.tools)[self.tool]\n                kwargs = {}\n                for arg in tool_args:\n                    value = getattr(self, arg.name, None)\n                    if value is not None:\n                        if isinstance(value, Message):\n                            kwargs[arg.name] = value.text\n                        else:\n                            kwargs[arg.name] = value\n\n                unflattened_kwargs = maybe_unflatten_dict(kwargs)\n\n                output = await exec_tool.coroutine(**unflattened_kwargs)\n\n                tool_content = []\n                for item in output.content:\n                    item_dict = item.model_dump()\n                    tool_content.append(item_dict)\n                return DataFrame(data=tool_content)\n            return DataFrame(data=[{\"error\": \"You must select a tool\"}])\n        except Exception as e:\n            msg = f\"Error in build_output: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n\n    def _get_session_context(self) -> str | None:\n        \"\"\"Get the Langflow session ID for MCP session caching.\"\"\"\n        # Try to get session ID from the component's execution context\n        if hasattr(self, \"graph\") and hasattr(self.graph, \"session_id\"):\n            session_id = self.graph.session_id\n            # Include server name to ensure different servers get different sessions\n            server_name = \"\"\n            mcp_server = getattr(self, \"mcp_server\", None)\n            if isinstance(mcp_server, dict):\n                server_name = mcp_server.get(\"name\", \"\")\n            elif mcp_server:\n                server_name = str(mcp_server)\n            return f\"{session_id}_{server_name}\" if session_id else None\n        return None\n\n    async def _get_tools(self):\n        \"\"\"Get cached tools or update if necessary.\"\"\"\n        mcp_server = getattr(self, \"mcp_server\", None)\n        if not self._not_load_actions:\n            tools, _ = await self.update_tool_list(mcp_server)\n            return tools\n        return []\n"
+                "value": "from __future__ import annotations\n\nimport asyncio\nimport json\nimport uuid\nfrom typing import Any\n\nfrom langchain_core.tools import StructuredTool  # noqa: TC002\n\nfrom lfx.base.agents.utils import maybe_unflatten_dict, safe_cache_get, safe_cache_set\nfrom lfx.base.mcp.util import (\n    MCPStdioClient,\n    MCPStreamableHttpClient,\n    create_input_schema_from_json_schema,\n    update_tools,\n)\nfrom lfx.custom.custom_component.component_with_cache import ComponentWithCache\nfrom lfx.inputs.inputs import InputTypes  # noqa: TC001\nfrom lfx.io import BoolInput, DropdownInput, McpInput, MessageTextInput, Output\nfrom lfx.io.schema import flatten_schema, schema_to_langflow_inputs\nfrom lfx.log.logger import logger\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.message import Message\nfrom lfx.services.deps import get_settings_service, get_storage_service, session_scope\n\n\nclass MCPToolsComponent(ComponentWithCache):\n    schema_inputs: list = []\n    tools: list[StructuredTool] = []\n    _not_load_actions: bool = False\n    _tool_cache: dict = {}\n    _last_selected_server: str | None = None  # Cache for the last selected server\n\n    def __init__(self, **data) -> None:\n        super().__init__(**data)\n        # Initialize cache keys to avoid CacheMiss when accessing them\n        self._ensure_cache_structure()\n\n        # Initialize clients with access to the component cache\n        self.stdio_client: MCPStdioClient = MCPStdioClient(component_cache=self._shared_component_cache)\n        self.streamable_http_client: MCPStreamableHttpClient = MCPStreamableHttpClient(\n            component_cache=self._shared_component_cache\n        )\n\n    def _ensure_cache_structure(self):\n        \"\"\"Ensure the cache has the required structure.\"\"\"\n        # Check if servers key exists and is not CacheMiss\n        servers_value = safe_cache_get(self._shared_component_cache, \"servers\")\n        if servers_value is None:\n            safe_cache_set(self._shared_component_cache, \"servers\", {})\n\n        # Check if last_selected_server key exists and is not CacheMiss\n        last_server_value = safe_cache_get(self._shared_component_cache, \"last_selected_server\")\n        if last_server_value is None:\n            safe_cache_set(self._shared_component_cache, \"last_selected_server\", \"\")\n\n    default_keys: list[str] = [\n        \"code\",\n        \"_type\",\n        \"tool_mode\",\n        \"tool_placeholder\",\n        \"mcp_server\",\n        \"tool\",\n        \"use_cache\",\n        \"verify_ssl\",\n    ]\n\n    display_name = \"MCP Tools\"\n    description = \"Connect to an MCP server to use its tools.\"\n    documentation: str = \"https://docs.langflow.org/mcp-client\"\n    icon = \"Mcp\"\n    name = \"MCPTools\"\n\n    inputs = [\n        McpInput(\n            name=\"mcp_server\",\n            display_name=\"MCP Server\",\n            info=\"Select the MCP Server that will be used by this component\",\n            real_time_refresh=True,\n        ),\n        BoolInput(\n            name=\"use_cache\",\n            display_name=\"Use Cached Server\",\n            info=(\n                \"Enable caching of MCP Server and tools to improve performance. \"\n                \"Disable to always fetch fresh tools and server updates.\"\n            ),\n            value=False,\n            advanced=True,\n        ),\n        BoolInput(\n            name=\"verify_ssl\",\n            display_name=\"Verify SSL Certificate\",\n            info=(\n                \"Enable SSL certificate verification for HTTPS connections. \"\n                \"Disable only for development/testing with self-signed certificates.\"\n            ),\n            value=True,\n            advanced=True,\n        ),\n        DropdownInput(\n            name=\"tool\",\n            display_name=\"Tool\",\n            options=[],\n            value=\"\",\n            info=\"Select the tool to execute\",\n            show=False,\n            required=True,\n            real_time_refresh=True,\n        ),\n        MessageTextInput(\n            name=\"tool_placeholder\",\n            display_name=\"Tool Placeholder\",\n            info=\"Placeholder for the tool\",\n            value=\"\",\n            show=False,\n            tool_mode=False,\n        ),\n    ]\n\n    outputs = [\n        Output(display_name=\"Response\", name=\"response\", method=\"build_output\"),\n    ]\n\n    async def _validate_schema_inputs(self, tool_obj) -> list[InputTypes]:\n        \"\"\"Validate and process schema inputs for a tool.\"\"\"\n        try:\n            if not tool_obj or not hasattr(tool_obj, \"args_schema\"):\n                msg = \"Invalid tool object or missing input schema\"\n                raise ValueError(msg)\n\n            flat_schema = flatten_schema(tool_obj.args_schema.schema())\n            input_schema = create_input_schema_from_json_schema(flat_schema)\n            if not input_schema:\n                msg = f\"Empty input schema for tool '{tool_obj.name}'\"\n                raise ValueError(msg)\n\n            schema_inputs = schema_to_langflow_inputs(input_schema)\n            if not schema_inputs:\n                msg = f\"No input parameters defined for tool '{tool_obj.name}'\"\n                await logger.awarning(msg)\n                return []\n\n        except Exception as e:\n            msg = f\"Error validating schema inputs: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return schema_inputs\n\n    async def update_tool_list(self, mcp_server_value=None):\n        # Accepts mcp_server_value as dict {name, config} or uses self.mcp_server\n        mcp_server = mcp_server_value if mcp_server_value is not None else getattr(self, \"mcp_server\", None)\n        server_name = None\n        server_config_from_value = None\n        if isinstance(mcp_server, dict):\n            server_name = mcp_server.get(\"name\")\n            server_config_from_value = mcp_server.get(\"config\")\n        else:\n            server_name = mcp_server\n        if not server_name:\n            self.tools = []\n            return [], {\"name\": server_name, \"config\": server_config_from_value}\n\n        # Check if caching is enabled, default to False\n        use_cache = getattr(self, \"use_cache\", False)\n\n        # Use shared cache if available and caching is enabled\n        cached = None\n        if use_cache:\n            servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n            cached = servers_cache.get(server_name) if isinstance(servers_cache, dict) else None\n\n        if cached is not None:\n            try:\n                self.tools = cached[\"tools\"]\n                self.tool_names = cached[\"tool_names\"]\n                self._tool_cache = cached[\"tool_cache\"]\n                server_config_from_value = cached[\"config\"]\n            except (TypeError, KeyError, AttributeError) as e:\n                # Handle corrupted cache data by clearing it and continuing to fetch fresh tools\n                msg = f\"Unable to use cached data for MCP Server{server_name}: {e}\"\n                await logger.awarning(msg)\n                # Clear the corrupted cache entry\n                current_servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                if isinstance(current_servers_cache, dict) and server_name in current_servers_cache:\n                    current_servers_cache.pop(server_name)\n                    safe_cache_set(self._shared_component_cache, \"servers\", current_servers_cache)\n            else:\n                return self.tools, {\"name\": server_name, \"config\": server_config_from_value}\n\n        try:\n            try:\n                from langflow.api.v2.mcp import get_server\n                from langflow.services.database.models.user.crud import get_user_by_id\n            except ImportError as e:\n                msg = (\n                    \"Langflow MCP server functionality is not available. \"\n                    \"This feature requires the full Langflow installation.\"\n                )\n                raise ImportError(msg) from e\n            async with session_scope() as db:\n                if not self.user_id:\n                    msg = \"User ID is required for fetching MCP tools.\"\n                    raise ValueError(msg)\n                current_user = await get_user_by_id(db, self.user_id)\n\n                # Try to get server config from DB/API\n                server_config = await get_server(\n                    server_name,\n                    current_user,\n                    db,\n                    storage_service=get_storage_service(),\n                    settings_service=get_settings_service(),\n                )\n\n            # If get_server returns empty but we have a config, use it\n            if not server_config and server_config_from_value:\n                server_config = server_config_from_value\n\n            if not server_config:\n                self.tools = []\n                return [], {\"name\": server_name, \"config\": server_config}\n\n            # Add verify_ssl option to server config if not present\n            if \"verify_ssl\" not in server_config:\n                verify_ssl = getattr(self, \"verify_ssl\", True)\n                server_config[\"verify_ssl\"] = verify_ssl\n\n            _, tool_list, tool_cache = await update_tools(\n                server_name=server_name,\n                server_config=server_config,\n                mcp_stdio_client=self.stdio_client,\n                mcp_streamable_http_client=self.streamable_http_client,\n            )\n\n            self.tool_names = [tool.name for tool in tool_list if hasattr(tool, \"name\")]\n            self._tool_cache = tool_cache\n            self.tools = tool_list\n\n            # Cache the result only if caching is enabled\n            if use_cache:\n                cache_data = {\n                    \"tools\": tool_list,\n                    \"tool_names\": self.tool_names,\n                    \"tool_cache\": tool_cache,\n                    \"config\": server_config,\n                }\n\n                # Safely update the servers cache\n                current_servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                if isinstance(current_servers_cache, dict):\n                    current_servers_cache[server_name] = cache_data\n                    safe_cache_set(self._shared_component_cache, \"servers\", current_servers_cache)\n\n        except (TimeoutError, asyncio.TimeoutError) as e:\n            msg = f\"Timeout updating tool list: {e!s}\"\n            await logger.aexception(msg)\n            raise TimeoutError(msg) from e\n        except Exception as e:\n            msg = f\"Error updating tool list: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return tool_list, {\"name\": server_name, \"config\": server_config}\n\n    async def update_build_config(self, build_config: dict, field_value: str, field_name: str | None = None) -> dict:\n        \"\"\"Toggle the visibility of connection-specific fields based on the selected mode.\"\"\"\n        try:\n            if field_name == \"tool\":\n                try:\n                    if len(self.tools) == 0:\n                        try:\n                            self.tools, build_config[\"mcp_server\"][\"value\"] = await self.update_tool_list()\n                            build_config[\"tool\"][\"options\"] = [tool.name for tool in self.tools]\n                            build_config[\"tool\"][\"placeholder\"] = \"Select a tool\"\n                        except (TimeoutError, asyncio.TimeoutError) as e:\n                            msg = f\"Timeout updating tool list: {e!s}\"\n                            await logger.aexception(msg)\n                            if not build_config[\"tools_metadata\"][\"show\"]:\n                                build_config[\"tool\"][\"show\"] = True\n                                build_config[\"tool\"][\"options\"] = []\n                                build_config[\"tool\"][\"value\"] = \"\"\n                                build_config[\"tool\"][\"placeholder\"] = \"Timeout on MCP server\"\n                            else:\n                                build_config[\"tool\"][\"show\"] = False\n                        except ValueError:\n                            if not build_config[\"tools_metadata\"][\"show\"]:\n                                build_config[\"tool\"][\"show\"] = True\n                                build_config[\"tool\"][\"options\"] = []\n                                build_config[\"tool\"][\"value\"] = \"\"\n                                build_config[\"tool\"][\"placeholder\"] = \"Error on MCP Server\"\n                            else:\n                                build_config[\"tool\"][\"show\"] = False\n\n                    if field_value == \"\":\n                        return build_config\n                    tool_obj = None\n                    for tool in self.tools:\n                        if tool.name == field_value:\n                            tool_obj = tool\n                            break\n                    if tool_obj is None:\n                        msg = f\"Tool {field_value} not found in available tools: {self.tools}\"\n                        await logger.awarning(msg)\n                        return build_config\n                    await self._update_tool_config(build_config, field_value)\n                except Exception as e:\n                    build_config[\"tool\"][\"options\"] = []\n                    msg = f\"Failed to update tools: {e!s}\"\n                    raise ValueError(msg) from e\n                else:\n                    return build_config\n            elif field_name == \"mcp_server\":\n                if not field_value:\n                    build_config[\"tool\"][\"show\"] = False\n                    build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"value\"] = \"\"\n                    build_config[\"tool\"][\"placeholder\"] = \"\"\n                    build_config[\"tool_placeholder\"][\"tool_mode\"] = False\n                    self.remove_non_default_keys(build_config)\n                    return build_config\n\n                build_config[\"tool_placeholder\"][\"tool_mode\"] = True\n\n                current_server_name = field_value.get(\"name\") if isinstance(field_value, dict) else field_value\n                _last_selected_server = safe_cache_get(self._shared_component_cache, \"last_selected_server\", \"\")\n\n                # To avoid unnecessary updates, only proceed if the server has actually changed\n                if (_last_selected_server in (current_server_name, \"\")) and build_config[\"tool\"][\"show\"]:\n                    if current_server_name:\n                        servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                        if isinstance(servers_cache, dict):\n                            cached = servers_cache.get(current_server_name)\n                            if cached is not None and cached.get(\"tool_names\"):\n                                cached_tools = cached[\"tool_names\"]\n                                current_tools = build_config[\"tool\"][\"options\"]\n                                if current_tools == cached_tools:\n                                    return build_config\n                    else:\n                        return build_config\n\n                # Determine if \"Tool Mode\" is active by checking if the tool dropdown is hidden.\n                is_in_tool_mode = build_config[\"tools_metadata\"][\"show\"]\n                safe_cache_set(self._shared_component_cache, \"last_selected_server\", current_server_name)\n\n                # Check if tools are already cached for this server before clearing\n                cached_tools = None\n                if current_server_name:\n                    use_cache = getattr(self, \"use_cache\", True)\n                    if use_cache:\n                        servers_cache = safe_cache_get(self._shared_component_cache, \"servers\", {})\n                        if isinstance(servers_cache, dict):\n                            cached = servers_cache.get(current_server_name)\n                            if cached is not None:\n                                try:\n                                    cached_tools = cached[\"tools\"]\n                                    self.tools = cached_tools\n                                    self.tool_names = cached[\"tool_names\"]\n                                    self._tool_cache = cached[\"tool_cache\"]\n                                except (TypeError, KeyError, AttributeError) as e:\n                                    # Handle corrupted cache data by ignoring it\n                                    msg = f\"Unable to use cached data for MCP Server,{current_server_name}: {e}\"\n                                    await logger.awarning(msg)\n                                    cached_tools = None\n\n                # Only clear tools if we don't have cached tools for the current server\n                if not cached_tools:\n                    self.tools = []  # Clear previous tools only if no cache\n\n                self.remove_non_default_keys(build_config)  # Clear previous tool inputs\n\n                # Only show the tool dropdown if not in tool_mode\n                if not is_in_tool_mode:\n                    build_config[\"tool\"][\"show\"] = True\n                    if cached_tools:\n                        # Use cached tools to populate options immediately\n                        build_config[\"tool\"][\"options\"] = [tool.name for tool in cached_tools]\n                        build_config[\"tool\"][\"placeholder\"] = \"Select a tool\"\n                    else:\n                        # Show loading state only when we need to fetch tools\n                        build_config[\"tool\"][\"placeholder\"] = \"Loading tools...\"\n                        build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"value\"] = uuid.uuid4()\n                else:\n                    # Keep the tool dropdown hidden if in tool_mode\n                    self._not_load_actions = True\n                    build_config[\"tool\"][\"show\"] = False\n\n            elif field_name == \"tool_mode\":\n                build_config[\"tool\"][\"placeholder\"] = \"\"\n                build_config[\"tool\"][\"show\"] = not bool(field_value) and bool(build_config[\"mcp_server\"])\n                self.remove_non_default_keys(build_config)\n                self.tool = build_config[\"tool\"][\"value\"]\n                if field_value:\n                    self._not_load_actions = True\n                else:\n                    build_config[\"tool\"][\"value\"] = uuid.uuid4()\n                    build_config[\"tool\"][\"options\"] = []\n                    build_config[\"tool\"][\"show\"] = True\n                    build_config[\"tool\"][\"placeholder\"] = \"Loading tools...\"\n            elif field_name == \"tools_metadata\":\n                self._not_load_actions = False\n\n        except Exception as e:\n            msg = f\"Error in update_build_config: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n        else:\n            return build_config\n\n    def get_inputs_for_all_tools(self, tools: list) -> dict:\n        \"\"\"Get input schemas for all tools.\"\"\"\n        inputs = {}\n        for tool in tools:\n            if not tool or not hasattr(tool, \"name\"):\n                continue\n            try:\n                flat_schema = flatten_schema(tool.args_schema.schema())\n                input_schema = create_input_schema_from_json_schema(flat_schema)\n                langflow_inputs = schema_to_langflow_inputs(input_schema)\n                inputs[tool.name] = langflow_inputs\n            except (AttributeError, ValueError, TypeError, KeyError) as e:\n                msg = f\"Error getting inputs for tool {getattr(tool, 'name', 'unknown')}: {e!s}\"\n                logger.exception(msg)\n                continue\n        return inputs\n\n    def remove_input_schema_from_build_config(\n        self, build_config: dict, tool_name: str, input_schema: dict[list[InputTypes], Any]\n    ):\n        \"\"\"Remove the input schema for the tool from the build config.\"\"\"\n        # Keep only schemas that don't belong to the current tool\n        input_schema = {k: v for k, v in input_schema.items() if k != tool_name}\n        # Remove all inputs from other tools\n        for value in input_schema.values():\n            for _input in value:\n                if _input.name in build_config:\n                    build_config.pop(_input.name)\n\n    def remove_non_default_keys(self, build_config: dict) -> None:\n        \"\"\"Remove non-default keys from the build config.\"\"\"\n        for key in list(build_config.keys()):\n            if key not in self.default_keys:\n                build_config.pop(key)\n\n    async def _update_tool_config(self, build_config: dict, tool_name: str) -> None:\n        \"\"\"Update tool configuration with proper error handling.\"\"\"\n        if not self.tools:\n            self.tools, build_config[\"mcp_server\"][\"value\"] = await self.update_tool_list()\n\n        if not tool_name:\n            return\n\n        tool_obj = next((tool for tool in self.tools if tool.name == tool_name), None)\n        if not tool_obj:\n            msg = f\"Tool {tool_name} not found in available tools: {self.tools}\"\n            self.remove_non_default_keys(build_config)\n            build_config[\"tool\"][\"value\"] = \"\"\n            await logger.awarning(msg)\n            return\n\n        try:\n            # Store current values before removing inputs\n            current_values = {}\n            for key, value in build_config.items():\n                if key not in self.default_keys and isinstance(value, dict) and \"value\" in value:\n                    current_values[key] = value[\"value\"]\n\n            # Get all tool inputs and remove old ones\n            input_schema_for_all_tools = self.get_inputs_for_all_tools(self.tools)\n            self.remove_input_schema_from_build_config(build_config, tool_name, input_schema_for_all_tools)\n\n            # Get and validate new inputs\n            self.schema_inputs = await self._validate_schema_inputs(tool_obj)\n            if not self.schema_inputs:\n                msg = f\"No input parameters to configure for tool '{tool_name}'\"\n                await logger.ainfo(msg)\n                return\n\n            # Add new inputs to build config\n            for schema_input in self.schema_inputs:\n                if not schema_input or not hasattr(schema_input, \"name\"):\n                    msg = \"Invalid schema input detected, skipping\"\n                    await logger.awarning(msg)\n                    continue\n\n                try:\n                    name = schema_input.name\n                    input_dict = schema_input.to_dict()\n                    input_dict.setdefault(\"value\", None)\n                    input_dict.setdefault(\"required\", True)\n\n                    build_config[name] = input_dict\n\n                    # Preserve existing value if the parameter name exists in current_values\n                    if name in current_values:\n                        build_config[name][\"value\"] = current_values[name]\n\n                except (AttributeError, KeyError, TypeError) as e:\n                    msg = f\"Error processing schema input {schema_input}: {e!s}\"\n                    await logger.aexception(msg)\n                    continue\n        except ValueError as e:\n            msg = f\"Schema validation error for tool {tool_name}: {e!s}\"\n            await logger.aexception(msg)\n            self.schema_inputs = []\n            return\n        except (AttributeError, KeyError, TypeError) as e:\n            msg = f\"Error updating tool config: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n\n    async def build_output(self) -> DataFrame:\n        \"\"\"Build output with improved error handling and validation.\"\"\"\n        try:\n            self.tools, _ = await self.update_tool_list()\n            if self.tool != \"\":\n                # Set session context for persistent MCP sessions using Langflow session ID\n                session_context = self._get_session_context()\n                if session_context:\n                    self.stdio_client.set_session_context(session_context)\n                    self.streamable_http_client.set_session_context(session_context)\n                exec_tool = self._tool_cache[self.tool]\n                tool_args = self.get_inputs_for_all_tools(self.tools)[self.tool]\n                kwargs = {}\n                for arg in tool_args:\n                    value = getattr(self, arg.name, None)\n                    if value is not None:\n                        if isinstance(value, Message):\n                            kwargs[arg.name] = value.text\n                        else:\n                            kwargs[arg.name] = value\n\n                unflattened_kwargs = maybe_unflatten_dict(kwargs)\n\n                output = await exec_tool.coroutine(**unflattened_kwargs)\n                tool_content = []\n                for item in output.content:\n                    item_dict = item.model_dump()\n                    item_dict = self.process_output_item(item_dict)\n                    tool_content.append(item_dict)\n\n                if isinstance(tool_content, list) and all(isinstance(x, dict) for x in tool_content):\n                    return DataFrame(tool_content)\n                return DataFrame(data=tool_content)\n            return DataFrame(data=[{\"error\": \"You must select a tool\"}])\n        except Exception as e:\n            msg = f\"Error in build_output: {e!s}\"\n            await logger.aexception(msg)\n            raise ValueError(msg) from e\n\n    def process_output_item(self, item_dict):\n        \"\"\"Process the output of a tool.\"\"\"\n        if item_dict.get(\"type\") == \"text\":\n            text = item_dict.get(\"text\")\n            try:\n                return json.loads(text)\n                # convert it to dict\n            except json.JSONDecodeError:\n                return item_dict\n        return item_dict\n\n    def _get_session_context(self) -> str | None:\n        \"\"\"Get the Langflow session ID for MCP session caching.\"\"\"\n        # Try to get session ID from the component's execution context\n        if hasattr(self, \"graph\") and hasattr(self.graph, \"session_id\"):\n            session_id = self.graph.session_id\n            # Include server name to ensure different servers get different sessions\n            server_name = \"\"\n            mcp_server = getattr(self, \"mcp_server\", None)\n            if isinstance(mcp_server, dict):\n                server_name = mcp_server.get(\"name\", \"\")\n            elif mcp_server:\n                server_name = str(mcp_server)\n            return f\"{session_id}_{server_name}\" if session_id else None\n        return None\n\n    async def _get_tools(self):\n        \"\"\"Get cached tools or update if necessary.\"\"\"\n        mcp_server = getattr(self, \"mcp_server\", None)\n        if not self._not_load_actions:\n            tools, _ = await self.update_tool_list(mcp_server)\n            return tools\n        return []\n"
               },
               "mcp_server": {
                 "_input_type": "McpInput",

langflow/initial_setup/starter_projects/Pok/303/251dex Agent.json"

@@ -401,7 +401,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Portfolio Website Code Generator.json

@@ -335,7 +335,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",
@@ -2072,7 +2072,7 @@
             "legacy": false,
             "lf_version": "1.6.0",
             "metadata": {
-              "code_hash": "a5b1b04d9fcc",
+              "code_hash": "9ad8d8d8fc16",
               "dependencies": {
                 "dependencies": [
                   {
@@ -2143,7 +2143,7 @@
                 "show": true,
                 "title_case": false,
                 "type": "code",
-                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(list[output_model_], Field(description=f\"A list of {schema_name}.\")),  # type: ignore[valid-type]\n        )\n\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[output_model])\n        except NotImplementedError as exc:\n            msg = f\"{self.llm.__class__.__name__} does not support structured output.\"\n            raise TypeError(msg) from exc\n\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        result = get_chat_result(\n            runnable=llm_with_structured_output,\n            system_message=self.system_prompt,\n            input_value=self.input_value,\n            config=config_dict,\n        )\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response.model_dump() if isinstance(first_response, BaseModel) else first_response\n\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n"
+                "value": "from pydantic import BaseModel, Field, create_model\nfrom trustcall import create_extractor\n\nfrom lfx.base.models.chat_result import get_chat_result\nfrom lfx.custom.custom_component.component import Component\nfrom lfx.helpers.base_model import build_model_from_schema\nfrom lfx.io import (\n    HandleInput,\n    MessageTextInput,\n    MultilineInput,\n    Output,\n    TableInput,\n)\nfrom lfx.log.logger import logger\nfrom lfx.schema.data import Data\nfrom lfx.schema.dataframe import DataFrame\nfrom lfx.schema.table import EditMode\n\n\nclass StructuredOutputComponent(Component):\n    display_name = \"Structured Output\"\n    description = \"Uses an LLM to generate structured data. Ideal for extraction and consistency.\"\n    documentation: str = \"https://docs.langflow.org/components-processing#structured-output\"\n    name = \"StructuredOutput\"\n    icon = \"braces\"\n\n    inputs = [\n        HandleInput(\n            name=\"llm\",\n            display_name=\"Language Model\",\n            info=\"The language model to use to generate the structured output.\",\n            input_types=[\"LanguageModel\"],\n            required=True,\n        ),\n        MultilineInput(\n            name=\"input_value\",\n            display_name=\"Input Message\",\n            info=\"The input message to the language model.\",\n            tool_mode=True,\n            required=True,\n        ),\n        MultilineInput(\n            name=\"system_prompt\",\n            display_name=\"Format Instructions\",\n            info=\"The instructions to the language model for formatting the output.\",\n            value=(\n                \"You are an AI that extracts structured JSON objects from unstructured text. \"\n                \"Use a predefined schema with expected types (str, int, float, bool, dict). \"\n                \"Extract ALL relevant instances that match the schema - if multiple patterns exist, capture them all. \"\n                \"Fill missing or ambiguous values with defaults: null for missing values. \"\n                \"Remove exact duplicates but keep variations that have different field values. \"\n                \"Always return valid JSON in the expected format, never throw errors. \"\n                \"If multiple objects can be extracted, return them all in the structured format.\"\n            ),\n            required=True,\n            advanced=True,\n        ),\n        MessageTextInput(\n            name=\"schema_name\",\n            display_name=\"Schema Name\",\n            info=\"Provide a name for the output data schema.\",\n            advanced=True,\n        ),\n        TableInput(\n            name=\"output_schema\",\n            display_name=\"Output Schema\",\n            info=\"Define the structure and data types for the model's output.\",\n            required=True,\n            # TODO: remove deault value\n            table_schema=[\n                {\n                    \"name\": \"name\",\n                    \"display_name\": \"Name\",\n                    \"type\": \"str\",\n                    \"description\": \"Specify the name of the output field.\",\n                    \"default\": \"field\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n                {\n                    \"name\": \"description\",\n                    \"display_name\": \"Description\",\n                    \"type\": \"str\",\n                    \"description\": \"Describe the purpose of the output field.\",\n                    \"default\": \"description of field\",\n                    \"edit_mode\": EditMode.POPOVER,\n                },\n                {\n                    \"name\": \"type\",\n                    \"display_name\": \"Type\",\n                    \"type\": \"str\",\n                    \"edit_mode\": EditMode.INLINE,\n                    \"description\": (\"Indicate the data type of the output field (e.g., str, int, float, bool, dict).\"),\n                    \"options\": [\"str\", \"int\", \"float\", \"bool\", \"dict\"],\n                    \"default\": \"str\",\n                },\n                {\n                    \"name\": \"multiple\",\n                    \"display_name\": \"As List\",\n                    \"type\": \"boolean\",\n                    \"description\": \"Set to True if this output field should be a list of the specified type.\",\n                    \"default\": \"False\",\n                    \"edit_mode\": EditMode.INLINE,\n                },\n            ],\n            value=[\n                {\n                    \"name\": \"field\",\n                    \"description\": \"description of field\",\n                    \"type\": \"str\",\n                    \"multiple\": \"False\",\n                }\n            ],\n        ),\n    ]\n\n    outputs = [\n        Output(\n            name=\"structured_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_output\",\n        ),\n        Output(\n            name=\"dataframe_output\",\n            display_name=\"Structured Output\",\n            method=\"build_structured_dataframe\",\n        ),\n    ]\n\n    def build_structured_output_base(self):\n        schema_name = self.schema_name or \"OutputModel\"\n\n        if not hasattr(self.llm, \"with_structured_output\"):\n            msg = \"Language model does not support structured output.\"\n            raise TypeError(msg)\n        if not self.output_schema:\n            msg = \"Output schema cannot be empty\"\n            raise ValueError(msg)\n\n        output_model_ = build_model_from_schema(self.output_schema)\n        output_model = create_model(\n            schema_name,\n            __doc__=f\"A list of {schema_name}.\",\n            objects=(\n                list[output_model_],\n                Field(\n                    description=f\"A list of {schema_name}.\",  # type: ignore[valid-type]\n                    min_length=1,  # help ensure non-empty output\n                ),\n            ),\n        )\n        # Tracing config\n        config_dict = {\n            \"run_name\": self.display_name,\n            \"project_name\": self.get_project_name(),\n            \"callbacks\": self.get_langchain_callbacks(),\n        }\n        # Generate structured output using Trustcall first, then fallback to Langchain if it fails\n        result = self._extract_output_with_trustcall(output_model, config_dict)\n        if result is None:\n            result = self._extract_output_with_langchain(output_model, config_dict)\n\n        # OPTIMIZATION NOTE: Simplified processing based on trustcall response structure\n        # Handle non-dict responses (shouldn't happen with trustcall, but defensive)\n        if not isinstance(result, dict):\n            return result\n\n        # Extract first response and convert BaseModel to dict\n        responses = result.get(\"responses\", [])\n        if not responses:\n            return result\n\n        # Convert BaseModel to dict (creates the \"objects\" key)\n        first_response = responses[0]\n        structured_data = first_response\n        if isinstance(first_response, BaseModel):\n            structured_data = first_response.model_dump()\n        # Extract the objects array (guaranteed to exist due to our Pydantic model structure)\n        return structured_data.get(\"objects\", structured_data)\n\n    def build_structured_output(self) -> Data:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            return Data(data=output[0])\n        if len(output) > 1:\n            # Multiple outputs - wrap them in a results container\n            return Data(data={\"results\": output})\n        return Data()\n\n    def build_structured_dataframe(self) -> DataFrame:\n        output = self.build_structured_output_base()\n        if not isinstance(output, list) or not output:\n            # handle empty or unexpected type case\n            msg = \"No structured output returned\"\n            raise ValueError(msg)\n        if len(output) == 1:\n            # For single dictionary, wrap in a list to create DataFrame with one row\n            return DataFrame([output[0]])\n        if len(output) > 1:\n            # Multiple outputs - convert to DataFrame directly\n            return DataFrame(output)\n        return DataFrame()\n\n    def _extract_output_with_trustcall(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = create_extractor(self.llm, tools=[schema], tool_choice=schema.__name__)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n        except Exception as e:  # noqa: BLE001\n            logger.warning(\n                f\"Trustcall extraction failed, falling back to Langchain: {e} \"\n                \"(Note: This may not be an error—some models or configurations do not support tool calling. \"\n                \"Falling back is normal in such cases.)\"\n            )\n            return None\n        return result or None  # langchain fallback is used if error occurs or the result is empty\n\n    def _extract_output_with_langchain(self, schema: BaseModel, config_dict: dict) -> list[BaseModel] | None:\n        try:\n            llm_with_structured_output = self.llm.with_structured_output(schema)\n            result = get_chat_result(\n                runnable=llm_with_structured_output,\n                system_message=self.system_prompt,\n                input_value=self.input_value,\n                config=config_dict,\n            )\n            if isinstance(result, BaseModel):\n                result = result.model_dump()\n                result = result.get(\"objects\", result)\n        except Exception as fallback_error:\n            msg = (\n                f\"Model does not support tool calling (trustcall failed) \"\n                f\"and fallback with_structured_output also failed: {fallback_error}\"\n            )\n            raise ValueError(msg) from fallback_error\n\n        return result or None\n"
               },
               "input_value": {
                 "_input_type": "MultilineInput",

langflow/initial_setup/starter_projects/Price Deal Finder.json

@@ -425,7 +425,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Research Agent.json

@@ -1650,7 +1650,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",

langflow/initial_setup/starter_projects/Research Translation Loop.json

@@ -411,7 +411,7 @@
                   },
                   {
                     "name": "fastapi",
-                    "version": "0.119.0"
+                    "version": "0.119.1"
                   },
                   {
                     "name": "lfx",