From 941e8feeb7fd1b17741bf11670554adfeb75b06e Mon Sep 17 00:00:00 2001 From: Lukas Schaefer Date: Tue, 21 Jul 2026 11:15:22 -0400 Subject: [PATCH] docs(taskprocessing): Document improve, and multimodal task types Signed-off-by: Lukas Schaefer --- .../digging_deeper/task_processing.rst | 31 +++++++++++++++++++ 1 file changed, 31 insertions(+) diff --git a/developer_manual/digging_deeper/task_processing.rst b/developer_manual/digging_deeper/task_processing.rst index e95c8e84467..874f5067b57 100644 --- a/developer_manual/digging_deeper/task_processing.rst +++ b/developer_manual/digging_deeper/task_processing.rst @@ -49,6 +49,18 @@ The following built-in task types are available: * Output shape: * ``output``: ``Text`` * ``tool_calls``: ``Text`` A string containing a JSON array with ``{"name": string, "type": "tool_call", "id": string, "args": object}`` + * ``'core:text2text:multimodal-chatwithtools'``: This task allows chatting with the language model with tools calling support and file attachments. It is implemented by ``\OCP\TaskProcessing\TaskTypes\MultimodalChatWithTools`` + * Input shape: + * ``system_prompt``: ``Text`` + * ``input``: ``Text`` + * ``input_attachments``: ``ListOfFiles`` Files to send along with the chat message + * ``tool_message``: ``Text`` A string containing a JSON array of ``{"name": string, "content": string, "tool_call_id": string}`` + * ``history``: ``ListOfTexts`` Each list item is a JSON string with ``{"role": "human", "content": string}`` or ``{"role": "assistant", "content": string, (optional: "tool_calls": array<{"name": string, "type": "tool_call", "id": string, "args": object}>)}`` or ``{"role": "tool", "content": string, "name": string, "tool_call_id": string}`` + * ``tools``: ``Text`` The tools parameter should be a JSON array formatted according to the OpenAI API specifications: https://platform.openai.com/docs/api-reference/chat/create#chat-create-tools + * Output shape: + * ``output``: ``Text`` + * ``output_attachments``: ``ListOfFiles`` Files generated by the model + * ``tool_calls``: ``Text`` A string containing a JSON array with ``{"name": string, "type": "tool_call", "id": string, "args": object}`` * ``'core:contextagent:interaction'``: This task allows chatting with an agent. It is implemented by ``\OCP\TaskProcessing\TaskTypes\ContextAgentInteraction`` * Input shape: * ``input``: ``Text`` @@ -58,6 +70,19 @@ The following built-in task types are available: * ``output``: ``Text`` * ``conversation_token``: ``Text`` * ``actions``: ``Text`` + * ``'core:contextagent:multimodal-interaction'``: This task allows chatting with an agent using file attachments. It is implemented by ``\OCP\TaskProcessing\TaskTypes\MultimodalContextAgentInteraction`` + * Input shape: + * ``input``: ``Text`` + * ``input_attachments``: ``ListOfFiles`` Files to send along with the chat message + * ``confirmation``: ``Number`` Boolean integer indicating whether to confirm previously requested actions: 0 to reject or 1 to confirm. + * ``conversation_token``: ``Text`` Token representing the conversation + * ``memories``: ``ListOfTexts`` Memories to send to the agent + * Output shape: + * ``output``: ``Text`` + * ``output_attachments``: ``ListOfFiles`` Files generated by the agent + * ``conversation_token``: ``Text`` + * ``actions``: ``Text`` + * ``sources``: ``ListOfTexts`` Sources of the generated response * ``'core:text2text:formalization'``: This task will reformulate the passed input text to be more formal in tone. It is implemented by ``\OCP\TaskProcessing\TaskTypes\TextToTextFormalization`` * Input shape: * ``input``: ``Text`` @@ -68,6 +93,12 @@ The following built-in task types are available: * ``input``: ``Text`` * Output shape: * ``output``: ``Text`` + * ``'core:text2text:improve'``: This task will improve the passed input text based on instructions. It is implemented by ``\OCP\TaskProcessing\TaskTypes\TextToTextImprove`` + * Input shape: + * ``input``: ``Text`` + * ``instructions``: ``Text`` + * Output shape: + * ``output``: ``Text`` * ``'core:text2text:reformulation'``: This task will reformulate the passed input text arbitrarily. It is implemented by ``\OCP\TaskProcessing\TaskTypes\TextToTextReformulation`` * Input shape: * ``input``: ``Text``