Update agent framework integrations (#4090)

* Update agent framework integrations

* Update agent framework integrations

* Update agent framework integrations

* Update README and optimize agent execution script

* Refactor script to improve readability and maintainability

* Update agent framework README files with pipeline deployment support

- Add zenml init/login setup steps in separate blocks
- Update deployment commands to use run.agent_pipeline format
- Replace -service naming with -agent naming convention
- Add comprehensive deployment documentation sections
- Include both CLI and HTTP API invocation examples
- Maintain consistent structure across all framework examples

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

* Refactor exception handling for file deletion

---------

Co-authored-by: Claude <[email protected]>
Co-authored-by: Alex Strick van Linschoten <[email protected]>

Author: 3 people

examples/agent_framework_integrations/README.md

@@ -44,11 +44,34 @@ Choose any framework and get started in minutes:
 export OPENAI_API_KEY="your-api-key-here"
 cd framework-name/
 uv venv --python 3.11
-source .venv/bin/activate  
+source .venv/bin/activate
 uv pip install -r requirements.txt
 python run.py
 ```
 
+### 🌐 Pipeline Deployments
+
+All examples support **real-time deployment as HTTP services** using ZenML's [Pipeline Deployment feature](https://docs.zenml.io/concepts/deployment):
+
+```bash
+# Deploy any agent pipeline as a persistent HTTP service
+zenml pipeline deploy agent_pipeline --name my-agent-service
+
+# Invoke via CLI
+zenml deployment invoke my-agent-service --query="Your question here"
+
+# Invoke via HTTP API
+curl -X POST http://localhost:8000/invoke \
+  -H "Content-Type: application/json" \
+  -d '{"parameters": {"query": "Your question here"}}'
+```
+
+This transforms batch-style agent workflows into **real-time APIs** perfect for:
+- **Web applications**: Integrate agents directly into your apps
+- **Microservices**: Deploy agents as independent services
+- **Production inference**: Serve models with preprocessing pipelines
+- **Interactive demos**: Create shareable agent endpoints
+
 ## 📊 Frameworks Overview
 
 | Framework | Type | Key Features | Technologies |
@@ -78,16 +101,13 @@ All examples follow these established patterns:
 ### 📦 Pipeline Architecture
 ```python
 @pipeline
-def agent_pipeline() -> str:
-    # 1. External artifact input
-    query = ExternalArtifact(value="Your query")
-    
-    # 2. Agent execution
+def agent_pipeline(query: str = "Your default query") -> str:
+    # 1. Agent execution with parameterized input
     results = run_agent(query)
-    
-    # 3. Response formatting
+
+    # 2. Response formatting
     summary = format_response(results)
-    
+
     return summary
 ```
 
@@ -124,14 +144,14 @@ def agent_pipeline() -> str:
 
 ## 🔄 Implementation Notes
 
-### Production vs. Demos
-**These examples demonstrate single-query execution for simplicity.** In production, ZenML's value comes from:
+### Production Use Cases
+**These examples support both batch and real-time execution.** ZenML's value in production includes:
 - **Batch processing**: Process hundreds/thousands of queries overnight
-- **Agent evaluation**: Compare different frameworks on test datasets  
+- **Real-time serving**: Deploy agents as HTTP APIs for instant responses
+- **Agent evaluation**: Compare different frameworks on test datasets
 - **Data pipelines**: Use agents to process document collections
 - **A/B testing**: Systematic comparison of agent configurations
-
-For real-time serving, use FastAPI/Flask directly. Use ZenML for the operational layer.
+- **Microservices**: Independent, scalable agent deployments
 
 ### Async Frameworks
 Some frameworks require async handling within ZenML steps:
@@ -174,6 +194,8 @@ ZenML is an extensible, open-source MLOps framework for creating production-read
 - 🎯 **Production ready**: Built-in monitoring, logging, and error handling
 - 🔧 **Tool agnostic**: Works with any agent framework
 - ☁️ **Cloud native**: Deploy anywhere with consistent behavior
+- 🌐 **Real-time APIs**: Deploy agents as HTTP services instantly
+- 🚀 **Pipeline deployments**: Transform batch workflows into live services
 
 ## 📖 Learn More
 

examples/agent_framework_integrations/__init__.py

@@ -0,0 +1,15 @@
+# Copyright (c) ZenML GmbH 2025. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+#
+#      https://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
+# or implied. See the License for the specific language governing
+# permissions and limitations under the License.
+"""Agent framework integrations examples."""
+

examples/agent_framework_integrations/autogen/README.md

@@ -1,17 +1,48 @@
 # Autogen + ZenML
 
-Multi-agent conversation framework integrated with ZenML.
+Multi-agent conversation framework integrated with ZenML for travel planning with collaborative agents.
+
+## 🚀 Quick Run
 
-## Run
 ```bash
 export OPENAI_API_KEY="your-api-key-here"
 uv venv --python 3.11
 source .venv/bin/activate
 uv pip install -r requirements.txt
+```
+
+Initialize ZenML and login:
+```bash
+zenml init
+zenml login
+```
+
+Run the pipeline:
+```bash
 python run.py
 ```
 
-## Features
-- Multi-agent conversations with UserProxy and Assistant agents
-- Async runtime management with proper cleanup
-- Travel planning use case with collaborative agents
+## 🌐 Pipeline Deployment
+
+Deploy this agent as a real-time HTTP service:
+
+```bash
+# Deploy the pipeline as an HTTP service
+zenml pipeline deploy run.agent_pipeline --name autogen-agent
+
+# Invoke via CLI
+zenml deployment invoke autogen-agent --destination="Tokyo" --days=3
+
+# Invoke via HTTP API
+curl -X POST http://localhost:8000/invoke \
+  -H "Content-Type: application/json" \
+  -d '{"parameters": {"destination": "Tokyo", "days": 3}}'
+```
+
+## ✨ Features
+
+- **Multi-Agent Collaboration**: Weather specialist and travel advisor agents
+- **Async Runtime Management**: Proper agent lifecycle with cleanup
+- **Travel Planning**: Comprehensive itinerary generation with weather integration
+- **Real-time Deployment**: Deploy as HTTP API for instant responses
+- **ZenML Orchestration**: Full pipeline tracking and artifact management

examples/agent_framework_integrations/autogen/__init__.py

@@ -0,0 +1,15 @@
+# Copyright (c) ZenML GmbH 2025. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+#
+#      https://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
+# or implied. See the License for the specific language governing
+# permissions and limitations under the License.
+"""AutoGen agent example."""
+

examples/agent_framework_integrations/autogen/requirements.txt

@@ -1,3 +1,5 @@
-autogen-core>=0.4.0
+autogen-core>=0.7.5
 autogen-ext[openai]>=0.4.0
 zenml[local]
+jinja2>=3.1.0
+secure>=0.3.0

examples/agent_framework_integrations/autogen/run.py

@@ -19,7 +19,7 @@
 )
 from pydantic import BaseModel
 
-from zenml import ExternalArtifact, pipeline, step
+from zenml import pipeline, step
 from zenml.config import DockerSettings, PythonPackageInstaller
 
 docker_settings = DockerSettings(
@@ -32,29 +32,29 @@
 
 
 class TravelQuery(BaseModel):
-    """User wants to plan a trip"""
+    """User wants to plan a trip."""
 
     destination: str
     days: int
 
 
 class WeatherInfo(BaseModel):
-    """Weather information for a destination"""
+    """Weather information for a destination."""
 
     destination: str
     forecast: str
     temperature: str
 
 
 class AttractionInfo(BaseModel):
-    """Tourist attractions for a destination"""
+    """Tourist attractions for a destination."""
 
     destination: str
     attractions: List[str]
 
 
 class TravelPlan(BaseModel):
-    """Complete travel plan combining all information"""
+    """Complete travel plan combining all information."""
 
     destination: str
     days: int
@@ -67,16 +67,17 @@ class TravelPlan(BaseModel):
 
 
 class WeatherAgent(RoutedAgent):
-    """Expert in weather information"""
+    """Expert in weather information."""
 
     def __init__(self):
+        """Initialize the Weather Specialist agent."""
         super().__init__("Weather Specialist")
 
     @message_handler
     async def check_weather(
         self, message: TravelQuery, ctx: MessageContext
     ) -> WeatherInfo:
-        """Get weather for the destination"""
+        """Get weather for the destination."""
         # Simulate weather API call
         weather_data = {
             "Paris": ("Partly cloudy with chance of rain", "15°C"),
@@ -101,16 +102,17 @@ async def check_weather(
 
 
 class AttractionAgent(RoutedAgent):
-    """Expert in tourist attractions"""
+    """Expert in tourist attractions."""
 
     def __init__(self):
+        """Initialize the Tourism Specialist agent."""
         super().__init__("Tourism Specialist")
 
     @message_handler
     async def find_attractions(
         self, message: TravelQuery, ctx: MessageContext
     ) -> AttractionInfo:
-        """Find top attractions for the destination"""
+        """Find top attractions for the destination."""
         # Simulate attraction database
         attractions_db = {
             "Paris": [
@@ -153,16 +155,17 @@ class TravelCoordinator(RoutedAgent):
     """Coordinates between specialists to create a travel plan and returns it."""
 
     def __init__(self):
+        """Initialize the Travel Coordinator agent."""
         super().__init__("Travel Coordinator")
 
     @message_handler
     async def coordinate_travel_plan(
         self, message: TravelQuery, _
     ) -> TravelPlan:
-        """
-        Receives a query, fans out to specialist agents, gathers results,
-        and returns a complete travel plan. This is the main entrypoint
-        for an external request.
+        """Receive a query, fan out to specialist agents, and gather results.
+
+        This is the main entrypoint for an external request that creates
+        a complete travel plan.
         """
         print(f"📋 Coordinator: Planning trip to {message.destination}")
 
@@ -213,16 +216,10 @@ async def setup_runtime() -> AgentRuntime:
 
 @step
 def run_autogen_agents(
-    query: str,
+    destination: str,
+    days: int,
 ) -> Annotated[Dict[str, Any], "agent_results"]:
     """Execute the autogen multi-agent system and return results."""
-    # Parse destination and days from query string
-    # Format: "Plan a X-day trip to DESTINATION"
-    parts = query.split(" to ")
-    destination = parts[1] if len(parts) > 1 else "Unknown"
-    days_part = [w for w in query.split() if w.endswith("-day")][0]
-    days = int(days_part.replace("-day", ""))
-
     # Create TravelQuery for the agents
     travel_query = TravelQuery(destination=destination, days=days)
 
@@ -290,17 +287,14 @@ def format_travel_plan(
 
 
 @pipeline(settings={"docker": docker_settings}, enable_cache=False)
-def autogen_travel_pipeline() -> str:
+def agent_pipeline(destination: str = "Paris", days: int = 4) -> str:
     """ZenML pipeline that orchestrates the autogen multi-agent travel system.
 
     Returns:
         Formatted travel plan summary
     """
-    # External artifact for travel query
-    travel_query = ExternalArtifact(value="Plan a 4-day trip to Paris")
-
     # Run the autogen agents
-    plan_data = run_autogen_agents(travel_query)
+    plan_data = run_autogen_agents(destination=destination, days=days)
 
     # Format the results
     summary = format_travel_plan(plan_data)
@@ -310,6 +304,6 @@ def autogen_travel_pipeline() -> str:
 
 if __name__ == "__main__":
     print("🚀 Running autogen travel pipeline...")
-    run_result = autogen_travel_pipeline()
+    run_result = agent_pipeline()
     print("Pipeline completed successfully!")
     print("Check the ZenML dashboard for detailed results and artifacts.")