feat: WhatsApp interface V2 — media I/O, interactive tools, Team/Workflow support

cookbook/02_agents/10_human_in_the_loop/mixed_external_and_regular_tools.py

@@ -73,4 +73,3 @@ def get_user_location() -> str:
         )
 
     pprint.pprint_run_response(run_response)
-

cookbook/05_agent_os/interfaces/whatsapp/agent_with_media.py

@@ -39,10 +39,4 @@
 # ---------------------------------------------------------------------------
 
 if __name__ == "__main__":
-    """Run your AgentOS.
-
-    You can see the configuration and available apps at:
-    http://localhost:7777/config
-
-    """
     agent_os.serve(app="agent_with_media:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/agent_with_user_memory.py

@@ -48,7 +48,6 @@
         First introduce yourself and ask for their name then, ask about themeselves, their hobbies, what they like to do and what they like to talk about.
         Use DuckDuckGo search tool to find latest information about things in the conversations
     """),
-    debug_mode=True,
 )
 
 
@@ -65,10 +64,4 @@
 # ---------------------------------------------------------------------------
 
 if __name__ == "__main__":
-    """Run your AgentOS.
-
-    You can see the configuration and available apps at:
-    http://localhost:7777/config
-
-    """
     agent_os.serve(app="agent_with_user_memory:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/basic.py

@@ -40,10 +40,4 @@
 # ---------------------------------------------------------------------------
 
 if __name__ == "__main__":
-    """Run your AgentOS.
-
-    You can see the configuration and available apps at:
-    http://localhost:7777/config
-
-    """
     agent_os.serve(app="basic:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/image_generation_model.py

@@ -20,10 +20,9 @@
     id="image_generation_model",
     db=agent_db,
     model=Gemini(
-        id="models/gemini-2.5-flash-image",
+        id="gemini-3-pro-image-preview",
         response_modalities=["Text", "Image"],
     ),
-    debug_mode=True,
 )
 
 
@@ -40,10 +39,4 @@
 # ---------------------------------------------------------------------------
 
 if __name__ == "__main__":
-    """Run your AgentOS.
-
-    You can see the configuration and available apps at:
-    http://localhost:7777/config
-
-    """
     agent_os.serve(app="image_generation_model:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/interactive_concierge.py

@@ -0,0 +1,80 @@
+"""
+Interactive Concierge
+=====================
+
+A WhatsApp concierge that uses every interactive UI feature to help
+users find restaurants, activities, and entertainment.
+
+Showcases: reply buttons, list messages, location pins, reactions,
+mark-as-read, and image sending — all in one conversational flow.
+
+Requires:
+  WHATSAPP_ACCESS_TOKEN, WHATSAPP_PHONE_NUMBER_ID
+  ANTHROPIC_API_KEY
+  uvx (for geocode-mcp server)
+"""
+
+from agno.agent import Agent
+from agno.db.sqlite import SqliteDb
+from agno.models.anthropic import Claude
+from agno.os.app import AgentOS
+from agno.os.interfaces.whatsapp import Whatsapp
+from agno.tools.mcp import MCPTools
+from agno.tools.websearch import WebSearchTools
+from agno.tools.whatsapp import WhatsAppTools
+
+agent_db = SqliteDb(db_file="tmp/concierge.db")
+
+concierge_agent = Agent(
+    name="Concierge",
+    model=Claude(id="claude-sonnet-4-6"),
+    tools=[
+        WhatsAppTools(
+            enable_send_reply_buttons=True,
+            enable_send_list_message=True,
+            enable_send_location=True,
+            enable_send_reaction=True,
+            enable_mark_as_read=True,
+            enable_send_image=True,
+        ),
+        WebSearchTools(),
+        MCPTools(command="uvx --python 3.12 geocode-mcp"),
+    ],
+    db=agent_db,
+    instructions=[
+        "You are a friendly concierge that helps users find restaurants, activities, and entertainment.",
+        "Use WhatsApp interactive features to create a smooth, tap-driven experience.",
+        "Follow this flow:",
+        "1. Mark the incoming message as read using mark_as_read.",
+        "2. Greet the user and ask what they are in the mood for using send_reply_buttons "
+        "with options like: Dinner, Drinks, Entertainment.",
+        "3. When they pick, ask a follow-up preference using send_reply_buttons "
+        "(e.g., cuisine type for dinner, vibe for drinks).",
+        "4. Ask for their location or neighborhood (they can type it).",
+        "5. Search the web for matching venues in their area.",
+        "6. Present the top results using send_list_message with sections "
+        "(e.g., 'Top Picks' and 'Hidden Gems'), each row having a title and short description.",
+        "7. When they pick a venue from the list, use mcp_geocoding_get_coordinates to get "
+        "accurate latitude and longitude, then send the location using send_location.",
+        "8. Send an image of the venue if available using send_image with a URL from search results.",
+        "9. React to their original message with a contextual emoji using send_reaction.",
+        "Keep messages short and conversational. Use interactive elements instead of asking "
+        "the user to type whenever possible.",
+        "IMPORTANT: Do NOT send a long text summary that repeats what's already in an interactive message. "
+        "When sending reply buttons or list messages, only add a brief one-line intro — the interactive "
+        "element IS the message. Never use asterisks (*) around emoji for bold formatting.",
+    ],
+    add_history_to_context=True,
+    num_history_runs=10,
+    add_datetime_to_context=True,
+    markdown=True,
+)
+
+agent_os = AgentOS(
+    agents=[concierge_agent],
+    interfaces=[Whatsapp(agent=concierge_agent, send_user_number_to_context=True)],
+)
+app = agent_os.get_app()
+
+if __name__ == "__main__":
+    agent_os.serve(app="interactive_concierge:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/multimodal_team.py

@@ -0,0 +1,74 @@
+"""
+Multimodal Team
+===============
+
+A coordinated team with a Vision Analyst and a Creative Agent that
+handles image analysis, image generation, and web research over WhatsApp.
+
+Send a photo to get it analyzed, or ask for an image to be generated.
+For redesign requests, the team analyzes first then creates.
+
+Requires:
+  WHATSAPP_ACCESS_TOKEN, WHATSAPP_PHONE_NUMBER_ID
+  OPENAI_API_KEY
+"""
+
+from agno.agent import Agent
+from agno.models.openai import OpenAIChat
+from agno.os.app import AgentOS
+from agno.os.interfaces.whatsapp import Whatsapp
+from agno.team import Team
+from agno.tools.dalle import DalleTools
+from agno.tools.websearch import WebSearchTools
+
+vision_analyst = Agent(
+    name="Vision Analyst",
+    model=OpenAIChat(id="gpt-4o"),
+    role="Analyzes images, files, and visual content in detail.",
+    instructions=[
+        "You are an expert visual analyst.",
+        "When given an image, describe it thoroughly: subjects, colors, composition, text, mood.",
+        "When given files (CSV, code, text), analyze their content and provide insights.",
+        "Keep analysis concise but detailed.",
+    ],
+    markdown=True,
+)
+
+creative_agent = Agent(
+    name="Creative Agent",
+    model=OpenAIChat(id="gpt-4o"),
+    role="Generates images with DALL-E and searches the web.",
+    tools=[DalleTools(), WebSearchTools()],
+    instructions=[
+        "You are a creative assistant with image generation abilities.",
+        "Use DALL-E to generate images when asked.",
+        "Use web search when you need reference information.",
+        "Describe generated images briefly after creation.",
+    ],
+    markdown=True,
+)
+
+multimodal_team = Team(
+    name="Multimodal Team",
+    mode="coordinate",
+    model=OpenAIChat(id="gpt-4o"),
+    members=[vision_analyst, creative_agent],
+    instructions=[
+        "Route image analysis and file analysis tasks to Vision Analyst.",
+        "Route image generation and web search tasks to Creative Agent.",
+        "If the user sends an image and asks to recreate or modify it, "
+        "first ask Vision Analyst to describe it, then ask Creative Agent "
+        "to generate a new version based on that description.",
+    ],
+    show_members_responses=False,
+    markdown=True,
+)
+
+agent_os = AgentOS(
+    teams=[multimodal_team],
+    interfaces=[Whatsapp(team=multimodal_team)],
+)
+app = agent_os.get_app()
+
+if __name__ == "__main__":
+    agent_os.serve(app="multimodal_team:app", reload=True)

cookbook/05_agent_os/interfaces/whatsapp/multimodal_workflow.py

@@ -0,0 +1,102 @@
+"""
+Multimodal Workflow
+===================
+
+A parallel workflow that runs visual analysis and web research
+simultaneously, then synthesizes findings with optional image
+generation and PDF output — all delivered over WhatsApp.
+
+Workflow structure:
+  Parallel:
+    - Visual Analysis (analyzes input images/files)
+    - Web Research (searches for related context)
+  Sequential:
+    - Creative Synthesis (combines results, generates images/PDFs)
+
+Requires:
+  WHATSAPP_ACCESS_TOKEN, WHATSAPP_PHONE_NUMBER_ID
+  OPENAI_API_KEY
+"""
+
+from agno.agent import Agent
+from agno.models.openai import OpenAIChat
+from agno.os.app import AgentOS
+from agno.os.interfaces.whatsapp import Whatsapp
+from agno.tools.dalle import DalleTools
+from agno.tools.file_generation import FileGenerationTools
+from agno.tools.websearch import WebSearchTools
+from agno.workflow import Parallel, Step, Workflow
+
+analyst = Agent(
+    name="Visual Analyst",
+    model=OpenAIChat(id="gpt-4o"),
+    instructions=[
+        "Analyze any images or files provided.",
+        "Describe visual elements, composition, colors, mood.",
+        "If no image, analyze the text topic visually.",
+        "Keep analysis concise but detailed.",
+    ],
+    markdown=True,
+)
+
+researcher = Agent(
+    name="Web Researcher",
+    model=OpenAIChat(id="gpt-4o"),
+    tools=[WebSearchTools()],
+    instructions=[
+        "Search the web for information related to the user's request.",
+        "Provide relevant facts, trends, and context.",
+    ],
+    markdown=True,
+)
+
+synthesizer = Agent(
+    name="Creative Synthesizer",
+    model=OpenAIChat(id="gpt-4o"),
+    tools=[DalleTools(), FileGenerationTools()],
+    instructions=[
+        "Combine the analysis and research from previous steps.",
+        "If the user asked for an image, generate one with DALL-E.",
+        "If the user asked for a report or document, generate a PDF.",
+        "Provide a final comprehensive response.",
+    ],
+    markdown=True,
+)
+
+analysis_step = Step(
+    name="Visual Analysis",
+    agent=analyst,
+    description="Analyze input images/files or describe the topic visually",
+)
+
+research_step = Step(
+    name="Web Research",
+    agent=researcher,
+    description="Search the web for related context and information",
+)
+
+research_phase = Parallel(
+    analysis_step,
+    research_step,
+    name="Research Phase",
+)
+
+synthesis_step = Step(
+    name="Creative Synthesis",
+    agent=synthesizer,
+    description="Combine analysis + research into a final response, generate images or PDFs if requested",
+)
+
+creative_workflow = Workflow(
+    name="Creative Pipeline",
+    steps=[research_phase, synthesis_step],
+)
+
+agent_os = AgentOS(
+    workflows=[creative_workflow],
+    interfaces=[Whatsapp(workflow=creative_workflow)],
+)
+app = agent_os.get_app()
+
+if __name__ == "__main__":
+    agent_os.serve(app="multimodal_workflow:app", reload=True)