pygeai 0.6.0b11__py3-none-any.whl → 0.6.0b13__py3-none-any.whl

pygeai/_docs/source/content/api_reference/chat.rst

@@ -144,7 +144,7 @@ Using variables and thread ID:
     print(response)
 
 High Level Service Layer
-^^^^^^^^^^^^^^^^^^^^^^^
+^^^^^^^^^^^^^^^^^^^^^^^^
 
 The `ChatManager` class provides a high-level interface for generating chat completions. It does not support streaming responses but simplifies the process by using structured models like `ChatMessageList` and `LlmSettings`.
 
@@ -325,4 +325,430 @@ With tool choice:
         tool_choice=tool_choice,
         tools=tools
     )
-    print(response)
+    print(response)
+
+Image Generation
+~~~~~~~~~~~~~~~~
+
+The GEAI SDK provides functionality to generate images using AI models. This can be achieved through the command line interface or the low-level service layer (ChatClient).
+
+Command Line
+^^^^^^^^^^^^
+
+The `geai chat generate-image` command generates images based on the provided model and parameters.
+
+.. code-block:: shell
+
+    geai chat generate-image \
+      --model "openai/dall-e-3" \
+      --prompt "A futuristic city with flying cars at sunset" \
+      --n 1 \
+      --quality "hd" \
+      --size "1024x1024" \
+      --aspect-ratio "1:1"
+
+Generate multiple images with different aspect ratio:
+
+.. code-block:: shell
+
+    geai chat generate-image \
+      --model "openai/dall-e-3" \
+      --prompt "A serene mountain landscape with a lake" \
+      --n 2 \
+      --quality "standard" \
+      --size "1792x1024" \
+      --aspect-ratio "16:9"
+
+Low Level Service Layer
+^^^^^^^^^^^^^^^^^^^^^^^
+
+The `ChatClient.generate_image` method provides a low-level interface to generate images.
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.generate_image(
+        model="openai/dall-e-3",
+        prompt="A futuristic city with flying cars at sunset",
+        n=1,
+        quality="hd",
+        size="1024x1024",
+        aspect_ratio="1:1"
+    )
+    print(response)
+
+Generate images without aspect ratio specification:
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.generate_image(
+        model="openai/dall-e-3",
+        prompt="An abstract painting with vibrant colors",
+        n=1,
+        quality="standard",
+        size="1024x1024"
+    )
+    print(response)
+
+Image Editing
+~~~~~~~~~~~~~
+
+The GEAI SDK provides functionality to edit existing images using AI models. This can be achieved through the command line interface or the low-level service layer (ChatClient).
+
+Command Line
+^^^^^^^^^^^^
+
+The `geai chat edit-image` command edits an existing image based on the provided instructions.
+
+.. code-block:: shell
+
+    geai chat edit-image \
+      --model "openai/dall-e-2" \
+      --prompt "Remove the ball from the image" \
+      --image "https://example.com/image.jpg" \
+      --size "1024x1024" \
+      --n 1 \
+      --quality "high"
+
+Edit with multiple variations:
+
+.. code-block:: shell
+
+    geai chat edit-image \
+      --model "openai/dall-e-2" \
+      --prompt "Change the background to a beach scene" \
+      --image "https://example.com/photo.jpg" \
+      --size "512x512" \
+      --n 3 \
+      --quality "standard"
+
+Low Level Service Layer
+^^^^^^^^^^^^^^^^^^^^^^^
+
+The `ChatClient.edit_image` method provides a low-level interface to edit images.
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.edit_image(
+        model="openai/dall-e-2",
+        prompt="Remove the ball from the image",
+        image="https://example.com/image.jpg",
+        size="1024x1024",
+        n=1,
+        quality="high"
+    )
+    print(response)
+
+Edit with default parameters:
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.edit_image(
+        model="openai/dall-e-2",
+        prompt="Add a rainbow to the sky",
+        image="https://example.com/landscape.jpg",
+        size="512x512"
+    )
+    print(response)
+
+Responses API
+~~~~~~~~~~~~~
+
+The GEAI SDK provides a Responses API that supports processing images and PDF files alongside text input. This API is particularly useful for multi-modal applications that need to analyze documents or images.
+
+Command Line
+^^^^^^^^^^^^
+
+The `geai chat response` command generates a response using the Responses API with support for file uploads.
+
+.. code-block:: shell
+
+    geai chat response \
+      --model "openai/gpt-4o" \
+      --input "What do you see in this image?" \
+      --files '["image.jpg"]' \
+      --temperature 0.7 \
+      --max-output-tokens 1000
+
+Process multiple files with tools:
+
+.. code-block:: shell
+
+    geai chat response \
+      --model "openai/gpt-4o" \
+      --input "Analyze these documents and extract key information" \
+      --files '["doc1.pdf", "doc2.pdf", "chart.jpg"]' \
+      --tools '[{"name": "extract_data", "description": "Extracts structured data", "parameters": {"type": "object"}}]' \
+      --tool-choice "auto" \
+      --temperature 0.5 \
+      --max-output-tokens 2000 \
+      --stream 1
+
+With reasoning and metadata:
+
+.. code-block:: shell
+
+    geai chat response \
+      --model "openai/o1-pro" \
+      --input "Solve this mathematical problem shown in the image" \
+      --files '["math_problem.jpg"]' \
+      --instructions "Show your work step by step" \
+      --reasoning '{"effort": "high"}' \
+      --metadata '{"task": "math_solving", "user_id": "123"}' \
+      --temperature 1.0
+
+Low Level Service Layer
+^^^^^^^^^^^^^^^^^^^^^^^
+
+The `ChatClient.get_response` method provides a low-level interface for the Responses API.
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.get_response(
+        model="openai/gpt-4o",
+        input="What do you see in this image?",
+        files=["image.jpg"],
+        temperature=0.7,
+        max_output_tokens=1000
+    )
+    print(response)
+
+Streaming response with files:
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.get_response(
+        model="openai/gpt-4o",
+        input="Describe the content of these images",
+        files=["image1.jpg", "image2.jpg"],
+        stream=True,
+        temperature=0.7
+    )
+
+    for chunk in response:
+        print(chunk, end="")
+
+With tools and advanced options:
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    tools = [
+        {
+            "name": "analyze_document",
+            "description": "Analyzes document content",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "topic": {"type": "string", "description": "Main topic"}
+                }
+            }
+        }
+    ]
+
+    response = client.get_response(
+        model="openai/gpt-4o",
+        input="Analyze this PDF and extract key topics",
+        files=["document.pdf"],
+        tools=tools,
+        tool_choice="auto",
+        temperature=0.5,
+        max_output_tokens=2000,
+        parallel_tool_calls=True,
+        metadata={"task": "document_analysis"},
+        user="user123"
+    )
+    print(response)
+
+With reasoning configuration:
+
+.. code-block:: python
+
+    from pygeai.chat.clients import ChatClient
+
+    client = ChatClient()
+
+    response = client.get_response(
+        model="openai/o1-pro",
+        input="Solve this complex problem from the image",
+        files=["problem.jpg"],
+        instructions="Think step by step and show your reasoning",
+        reasoning={"effort": "medium"},
+        temperature=1.0,
+        truncation="disabled",
+        store=True
+    )
+    print(response)
+
+Interactive Chat with Iris
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Iris is an AI assistant that helps guide users through the process of creating agents. The GEAI SDK provides an interactive chat interface for Iris.
+
+Command Line
+^^^^^^^^^^^^
+
+The `geai chat iris` command starts an interactive chat session with Iris.
+
+.. code-block:: shell
+
+    geai chat iris
+
+This opens an interactive prompt where you can ask Iris questions about creating and configuring agents. Type 'exit' or press Ctrl+C to close the chat.
+
+Python API
+^^^^^^^^^^
+
+The `Iris` class provides a programmatic interface to chat with Iris.
+
+.. code-block:: python
+
+    from pygeai.chat.iris import Iris
+
+    iris = Iris()
+    messages = []
+
+    user_message = {
+        "role": "user",
+        "content": "I want to create an agent for customer support"
+    }
+    messages.append(user_message)
+
+    result = iris.stream_answer(messages)
+    answer = ""
+    for chunk in result:
+        answer += chunk
+        print(chunk, end="")
+
+    assistant_message = {
+        "role": "assistant",
+        "content": answer
+    }
+    messages.append(assistant_message)
+
+Interactive Chat with Agents
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The GEAI SDK provides functionality to have interactive chat sessions with agents. This includes both command-line and GUI interfaces, as well as session management capabilities.
+
+Command Line
+^^^^^^^^^^^^
+
+The `geai chat agent` command starts an interactive chat session with a specified agent.
+
+Basic usage:
+
+.. code-block:: shell
+
+    geai chat agent --agent-name "my-support-agent"
+
+This opens an interactive prompt where you can chat with the agent. Type 'exit' or press Ctrl+C to close the chat.
+
+With GUI interface:
+
+.. code-block:: shell
+
+    geai chat agent --agent-name "my-support-agent" --gui
+
+This launches a Streamlit-based graphical user interface for chatting with the agent. The GUI provides a more user-friendly experience with a web-based interface.
+
+Note: The `--gui` option requires Streamlit to be installed. Install it with `pip install streamlit` if not already available.
+
+Save conversation to a file:
+
+.. code-block:: shell
+
+    geai chat agent --agent-name "my-support-agent" --save-session conversation.json
+
+This saves the conversation history to `conversation.json` after each message exchange. The session is automatically saved when you exit the chat.
+
+Restore a previous conversation:
+
+.. code-block:: shell
+
+    geai chat agent --agent-name "my-support-agent" --restore-session conversation.json
+
+This loads a previous conversation from `conversation.json` and continues the chat from where it left off. The conversation history is displayed before you can continue chatting.
+
+Combine session save and restore:
+
+.. code-block:: shell
+
+    geai chat agent --agent-name "my-support-agent" --restore-session old_conv.json --save-session new_conv.json
+
+This restores the conversation from `old_conv.json` and saves all subsequent messages to `new_conv.json`.
+
+Python API
+^^^^^^^^^^
+
+The `AgentChatSession` class provides a programmatic interface to chat with agents.
+
+.. code-block:: python
+
+    from pygeai.chat.session import AgentChatSession
+
+    session = AgentChatSession("my-support-agent")
+    messages = []
+
+    introduction = session.get_answer(
+        ["You're about to speak to a user. Introduce yourself in a clear and concise manner."]
+    )
+    print(f"Agent: {introduction}")
+
+    messages.append({"role": "assistant", "content": introduction})
+
+    user_message = {
+        "role": "user",
+        "content": "How can I reset my password?"
+    }
+    messages.append(user_message)
+
+    result = session.stream_answer(messages)
+    answer = ""
+    print("Agent: ", end="")
+    for chunk in result:
+        answer += chunk
+        print(chunk, end="")
+
+    messages.append({"role": "assistant", "content": answer})
+
+Non-streaming chat:
+
+.. code-block:: python
+
+    from pygeai.chat.session import AgentChatSession
+
+    session = AgentChatSession("my-support-agent")
+
+    user_messages = [
+        {"role": "user", "content": "What are your capabilities?"}
+    ]
+
+    answer = session.get_answer(user_messages)
+    print(f"Agent: {answer}")

pygeai/_docs/source/content/api_reference/embeddings.rst

@@ -7,7 +7,7 @@ The API Reference enables you to generate embeddings from various input types, i
 
 
 Generate embeddings
-~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~
 
 Generates embeddings from different input types using `PyGEA </pygeai>`_. It can interact with several LLM providers and their respective models for embedding generation.
 

pygeai/_docs/source/content/api_reference/project.rst

@@ -7,6 +7,7 @@ The API Reference provides a comprehensive set of tools for managing projects wi
 * Delete project: Deletes an existing project.
 * List projects: Retrieves a list of all projects within your organization. You can choose to retrieve a summary or detailed information for each project.
 * Update project: Updates the name and description of an existing project.
+* Add project member: Sends an invitation email to add a user to a project with specific roles (supports individual and batch invitations).
 
 List projects
 ~~~~~~~~~~~~~~
@@ -464,6 +465,189 @@ Replace `<project_id>` with the actual project Id. For example:
     print(f"project: {project}")
 
 
+Member Management
+=================
+You can manage project members to control access and permissions within your projects.
+
+* Add Project Member: Sends an invitation email to add a user to a project with specific roles.
+
+Add Project Member
+~~~~~~~~~~~~~~~~~~
+
+Sends an invitation email to add a user to a project in `Globant Enterprise AI <https://wiki.genexus.com/enterprise-ai/wiki?8,Table+of+contents%3AEnterprise+AI>`_ using `PyGEA </pygeai>`_. You can add individual users or process multiple invitations via a CSV file (batch mode).
+
+To achieve this, you have three options:
+
+* `Command Line </docs/source/content/api_reference.rst#command-line>`_
+* `Low-Level Service Layer </docs/source/content/api_reference.rst#low-level-service-layer>`_
+* `High-Level Service Layer </docs/source/content/api_reference.rst#high-level-service-layer>`_
+
+Command line
+^^^^^^^^^^^^
+
+Add a single user to a project
+###############################
+
+Use the following command to invite a user to a project:
+
+.. code-block:: shell
+
+    geai org add-project-member \
+      --project-id <project_id> \
+      --email <user_email> \
+      --roles "<role1>,<role2>"
+
+Replace the placeholders with the actual values:
+
+* `<project_id>`: Id of the project.
+* `<user_email>`: Email address of the user to invite.
+* `<role1>,<role2>`: Comma-separated list of roles (e.g., "Project member,Project administrator").
+
+For example:
+
+.. code-block:: shell
+
+    geai org add-project-member \
+      --project-id 1956c032-3c66-4435-acb8-6a06e52f819f \
+      --email user@example.com \
+      --roles "Project member,Project administrator"
+
+You can also use the short alias:
+
+.. code-block:: shell
+
+    geai org apm \
+      --project-id 1956c032-3c66-4435-acb8-6a06e52f819f \
+      --email user@example.com \
+      --roles "Project member"
+
+
+Add multiple users via CSV file (batch mode)
+#############################################
+
+Use the `-b` or `--batch` flag to process multiple invitations from a CSV file:
+
+.. code-block:: shell
+
+    geai org add-project-member --batch <path_to_csv_file>
+
+The CSV file format should be:
+
+.. code-block:: text
+
+    project_id,email,role1,role2,...
+    1956c032-3c66-4435-acb8-6a06e52f819f,user1@example.com,Project member
+    1956c032-3c66-4435-acb8-6a06e52f819f,user2@example.com,Project member,Project administrator
+
+For example:
+
+.. code-block:: shell
+
+    geai org add-project-member --batch project_members.csv
+
+The command will process each line and report successful and failed invitations.
+
+
+Low level service layer
+^^^^^^^^^^^^^^^^^^^^^^^
+
+Use the following code snippet to invite a user to a project using the low-level service layer:
+
+.. code-block:: python
+
+    from pygeai.organization.clients import OrganizationClient
+
+    project_id = "<project_id>"
+    email = "<user_email>"
+    roles = ["Project member", "Project administrator"]
+
+    client = OrganizationClient()
+    result = client.add_project_member(project_id=project_id, email=email, roles=roles)
+    print(result)
+
+Replace the placeholders with the actual values. For example:
+
+.. code-block:: python
+
+    from pygeai.organization.clients import OrganizationClient
+
+    project_id = "1956c032-3c66-4435-acb8-6a06e52f819f"
+    email = "user@example.com"
+    roles = ["Project member", "Project administrator"]
+
+    client = OrganizationClient()
+    result = client.add_project_member(project_id=project_id, email=email, roles=roles)
+    print(result)
+
+For batch processing, you can use:
+
+.. code-block:: python
+
+    import csv
+    from pygeai.organization.clients import OrganizationClient
+
+    client = OrganizationClient()
+
+    with open('project_members.csv', 'r') as file:
+        csv_reader = csv.reader(file)
+        for row in csv_reader:
+            if len(row) >= 3:
+                project_id = row[0]
+                email = row[1]
+                roles = row[2:]
+                try:
+                    result = client.add_project_member(
+                        project_id=project_id,
+                        email=email,
+                        roles=roles
+                    )
+                    print(f"✓ Invited {email} to project {project_id}")
+                except Exception as e:
+                    print(f"✗ Failed to invite {email}: {e}")
+
+
+High level service layer
+^^^^^^^^^^^^^^^^^^^^^^^^
+
+Use the following code snippet to invite a user to a project using the high-level service layer:
+
+.. code-block:: python
+
+    from pygeai.organization.managers import OrganizationManager
+
+    manager = OrganizationManager()
+
+    project_id = "<project_id>"
+    email = "<user_email>"
+    roles = ["Project member", "Project administrator"]
+
+    response = manager.add_project_member(
+        project_id=project_id,
+        email=email,
+        roles=roles
+    )
+    print(f"response: {response}")
+
+Replace the placeholders with the actual values. For example:
+
+.. code-block:: python
+
+    from pygeai.organization.managers import OrganizationManager
+
+    manager = OrganizationManager()
+
+    project_id = "1956c032-3c66-4435-acb8-6a06e52f819f"
+    email = "user@example.com"
+    roles = ["Project member", "Project administrator"]
+
+    response = manager.add_project_member(
+        project_id=project_id,
+        email=email,
+        roles=roles
+    )
+    print(f"response: {response}")
+
+
 Token Management
 ================
 You can manage project tokens, which are essential for authentication and authorization when interacting with specific projects.

pygeai/_docs/source/content/api_reference/rag.rst

@@ -604,7 +604,7 @@ High-Level Service Layer
 An ``EmptyResponse`` indicating success, or an error response.
 
 Delete All Documents
--------------------
+--------------------
 
 Deletes all documents associated with a RAG Assistant.
 
@@ -620,7 +620,7 @@ Command Line
 * ``-n, --name``: (Required) Name of the RAG Assistant.
 
 Low-Level Service Layer
-^^^^^^^^^^^^^^^^^^^^^^
+^^^^^^^^^^^^^^^^^^^^^^^
 
 .. code-block:: python