Add LangGraph Hello World sample

- Add langgraph_samples/basic/hello_world/ demonstrating basic
  Temporal + LangGraph integration with a single-node graph
- Add langgraph dependency group to pyproject.toml
- Update langchain dependency group to 0.3.x for compatibility
  with langgraph (requires langchain-core>=0.2.27)

langgraph_samples/README.md

@@ -0,0 +1,24 @@
+# Temporal LangGraph Samples
+
+These samples demonstrate the [Temporal LangGraph integration](https://github.com/temporalio/sdk-python/blob/main/temporalio/contrib/langgraph/README.md) - combining LangGraph's agent framework with Temporal's durable execution.
+
+See the [module documentation](https://github.com/temporalio/sdk-python/blob/main/temporalio/contrib/langgraph/README.md) for more information.
+
+## Overview
+
+The integration combines:
+- **Temporal workflows** for orchestrating agent control flow and state management
+- **LangGraph** for defining agent graphs with conditional logic, cycles, and state
+
+This approach ensures that AI agent workflows are durable, observable, and can handle failures gracefully.
+
+## Prerequisites
+
+- Temporal server [running locally](https://docs.temporal.io/cli/server#start-dev)
+- Required dependencies installed via `uv sync --group langgraph`
+
+## Examples
+
+Each directory contains complete examples with their own README for detailed instructions:
+
+- **[Basic Examples](./basic/README.md)** - Simple examples including a hello world agent demonstrating basic plugin setup and graph registration.

langgraph_samples/SAMPLES_PROPOSAL.md

@@ -0,0 +1,349 @@
+# Temporal LangGraph Samples Proposal
+
+This document proposes sample applications demonstrating the [Temporal LangGraph integration](https://github.com/temporalio/sdk-python/blob/main/temporalio/contrib/langgraph/README.md) - combining LangGraph's agent framework with Temporal's durable execution.
+
+## Design Principles
+
+Each sample should demonstrate:
+1. **Durability** - Agents that survive failures and can resume from checkpoints
+2. **Production Readiness** - Proper timeout/retry configuration, error handling
+3. **Temporal Features** - Leveraging signals, queries, child workflows where appropriate
+
+Samples are organized by complexity and use case, inspired by [OpenAI Agents samples](../openai_agents/README.md) structure.
+
+---
+
+## Proposed Samples
+
+### 1. Basic Examples (`basic/`)
+
+#### 1.1 Hello World Agent
+**Description:** Minimal LangGraph agent with Temporal integration. Single-node graph that processes a query.
+
+**Demonstrates:**
+- Basic plugin setup and graph registration
+- Simple workflow invocation
+- Activity-based node execution
+
+**Original Reference:** N/A (custom introductory example)
+
+---
+
+#### 1.2 ReAct Agent with Tools
+**Description:** Classic ReAct (Reasoning + Acting) agent that can use tools to answer questions. Implements the think-act-observe loop.
+
+**Demonstrates:**
+- `temporal_model()` for durable LLM calls
+- `temporal_tool()` for durable tool execution
+- `create_react_agent` integration
+- Automatic retries on LLM/tool failures
+
+**Original Reference:** [LangGraph ReAct Agent Template](https://github.com/langchain-ai/react-agent)
+
+---
+
+### 2. Human-in-the-Loop (`human_in_loop/`)
+
+#### 2.1 Approval Workflow
+**Description:** Agent that pauses for human approval before taking actions. Uses LangGraph's `interrupt()` with Temporal signals.
+
+**Demonstrates:**
+- `interrupt()` for pausing execution
+- Temporal signals for receiving human input
+- Workflow queries for checking pending approvals
+- Timeout handling for approval deadlines
+
+**Original Reference:** [LangGraph Human-in-the-Loop Guide](https://langchain-ai.github.io/langgraph/how-tos/human_in_the_loop/)
+
+---
+
+#### 2.2 Document Review Agent
+**Description:** Agent that drafts documents and requests human review/edits before finalizing. Supports iterative refinement cycles.
+
+**Demonstrates:**
+- Multi-step interrupt/resume cycles
+- State preservation across interrupts
+- Combining AI drafting with human editing
+
+**Original Reference:** [LangGraph Human Feedback Discussion](https://github.com/langchain-ai/langgraph/discussions/1137)
+
+---
+
+### 3. Multi-Agent Systems (`multi_agent/`)
+
+#### 3.1 Supervisor Agent
+**Description:** Supervisor that coordinates specialized worker agents (researcher, writer, critic). The supervisor routes tasks to appropriate agents and aggregates results.
+
+**Demonstrates:**
+- Multi-agent orchestration patterns
+- Supervisor decision-making logic
+- Agent handoffs and communication
+- Parallel agent execution with `Send`
+
+**Original Reference:** [LangGraph Supervisor Library](https://github.com/langchain-ai/langgraph-supervisor-py) | [Multi-Agent Blog Post](https://blog.langchain.com/langgraph-multi-agent-workflows/)
+
+---
+
+#### 3.2 Hierarchical Agent Teams
+**Description:** Two-level hierarchy with a top supervisor managing team supervisors (research team + writing team), each with their own worker agents.
+
+**Demonstrates:**
+- Child workflows for team isolation
+- Hierarchical agent coordination
+- Complex routing logic
+- Cross-team communication
+
+**Original Reference:** [LangGraph Hierarchical Agent Teams Tutorial](https://langchain-ai.github.io/langgraph/tutorials/multi_agent/hierarchical_agent_teams/)
+
+---
+
+### 4. RAG and Research (`rag/`)
+
+#### 4.1 Agentic RAG
+**Description:** RAG agent that decides when to retrieve documents vs. respond directly. Includes document relevance grading and query rewriting.
+
+**Demonstrates:**
+- Conditional retrieval decisions
+- Document grading nodes
+- Query rewriting for better retrieval
+- Combining retrieval with generation
+
+**Original Reference:** [LangGraph Agentic RAG Tutorial](https://github.com/langchain-ai/langgraph/blob/main/examples/rag/langgraph_agentic_rag.ipynb) | [Agentic RAG Docs](https://docs.langchain.com/oss/python/langgraph/agentic-rag)
+
+---
+
+#### 4.2 Corrective RAG
+**Description:** Self-correcting RAG that validates retrieved documents and falls back to web search when needed. Includes hallucination detection.
+
+**Demonstrates:**
+- Document relevance validation
+- Fallback to web search
+- Self-correction loops
+- Multiple retrieval strategies
+
+**Original Reference:** [Corrective RAG Guide](https://www.analyticsvidhya.com/blog/2024/07/building-agentic-rag-systems-with-langgraph/)
+
+---
+
+#### 4.3 Deep Research Agent
+**Description:** Multi-step research agent that performs iterative web searches, evaluates results, and synthesizes findings into a comprehensive report.
+
+**Demonstrates:**
+- Long-running research workflows (minutes to hours)
+- Parallel search execution
+- Result evaluation and iteration
+- Continue-as-new for large histories
+- Report generation
+
+**Original Reference:** [LangGraph Deep Research Agent](https://towardsdatascience.com/langgraph-101-lets-build-a-deep-research-agent/) | [Exa Research System](https://blog.langchain.com/exa/)
+
+---
+
+### 5. Customer Service (`customer_service/`)
+
+#### 5.1 Customer Support Agent
+**Description:** Interactive customer service agent that handles inquiries, accesses knowledge bases, and escalates to humans when needed.
+
+**Demonstrates:**
+- Conversational state management
+- Knowledge base integration
+- Escalation workflows with `interrupt()`
+- Session continuity across interactions
+
+**Original Reference:** [LangGraph Customer Support Tutorial](https://github.com/langchain-ai/langgraph/blob/main/docs/docs/tutorials/customer-support/customer-support.ipynb)
+
+---
+
+#### 5.2 Email Triage Agent
+**Description:** Agent that processes incoming emails, categorizes them, drafts responses using RAG for policy questions, and routes to appropriate handlers.
+
+**Demonstrates:**
+- Email classification
+- RAG for policy/knowledge lookup
+- Conditional routing based on email type
+- Draft review before sending
+
+**Original Reference:** [Agentic RAG Customer Support](https://blog.lancedb.com/agentic-rag-using-langgraph-building-a-simple-customer-support-autonomous-agent/)
+
+---
+
+### 6. Planning and Execution (`planning/`)
+
+#### 6.1 Plan-and-Execute Agent
+**Description:** Agent that first creates a plan, then executes steps sequentially, with the ability to replan based on results.
+
+**Demonstrates:**
+- Separation of planning and execution
+- Dynamic replanning
+- Step-by-step execution with checkpointing
+- Plan validation
+
+**Original Reference:** [LangGraph Plan-and-Execute Tutorial](https://github.com/langchain-ai/langgraph/blob/main/examples/plan-and-execute/plan-and-execute.ipynb)
+
+---
+
+#### 6.2 Travel Planning Agent
+**Description:** Agent that plans trips by coordinating flight searches, hotel bookings, and itinerary creation with user preferences.
+
+**Demonstrates:**
+- Multi-tool coordination
+- User preference handling
+- Complex planning with constraints
+- Real API integration patterns
+
+**Original Reference:** [LangGraph Travel Planning Tutorial](https://www.projectpro.io/article/langgraph/1109)
+
+---
+
+### 7. Code and Development (`code/`)
+
+#### 7.1 Code Assistant
+**Description:** Agent that helps with code-related tasks: explaining code, generating code, debugging, and answering programming questions.
+
+**Demonstrates:**
+- Code-aware prompting
+- Multi-turn coding conversations
+- Tool use for code execution/testing
+- Context management for large codebases
+
+**Original Reference:** [LangGraph Code Assistant Tutorial](https://github.com/langchain-ai/langgraph/blob/main/examples/code_assistant/langgraph_code_assistant.ipynb)
+
+---
+
+### 8. Advanced Patterns (`advanced/`)
+
+#### 8.1 Reflection Agent
+**Description:** Agent that generates output, reflects on it, and iteratively improves until quality criteria are met.
+
+**Demonstrates:**
+- Self-critique and improvement loops
+- Quality evaluation nodes
+- Iterative refinement patterns
+- Loop termination conditions
+
+**Original Reference:** [LangGraph Reflection Tutorial](https://github.com/langchain-ai/langgraph/tree/main/docs/docs/tutorials/reflection)
+
+---
+
+#### 8.2 Tree of Thoughts
+**Description:** Agent that explores multiple reasoning paths in parallel and selects the best solution.
+
+**Demonstrates:**
+- Parallel exploration with `Send`
+- Solution evaluation and ranking
+- Branch pruning logic
+- Complex graph patterns
+
+**Original Reference:** [LangGraph Tree of Thoughts Tutorial](https://github.com/langchain-ai/langgraph/tree/main/docs/docs/tutorials/tot)
+
+---
+
+#### 8.3 Long-Running Workflow with Checkpointing
+**Description:** Demonstrates continue-as-new for workflows that may run for extended periods and generate large event histories.
+
+**Demonstrates:**
+- `should_continue` callback for checkpointing
+- State serialization and restoration
+- Continue-as-new pattern
+- Store API for cross-workflow data
+
+**Original Reference:** N/A (Temporal-specific pattern)
+
+---
+
+## Implementation Priority
+
+### Phase 1 - Core Examples (Recommended First)
+1. **Basic: Hello World Agent** - Entry point for new users
+2. **Basic: ReAct Agent with Tools** - Most common agent pattern
+3. **Human-in-the-Loop: Approval Workflow** - Key differentiator for Temporal
+4. **RAG: Agentic RAG** - Popular production use case
+
+### Phase 2 - Multi-Agent and Research
+5. **Multi-Agent: Supervisor Agent** - Popular advanced pattern
+6. **RAG: Deep Research Agent** - Showcases long-running durability
+7. **Customer Service: Support Agent** - Business-relevant example
+
+### Phase 3 - Advanced Patterns
+8. **Planning: Plan-and-Execute** - Structured agent execution
+9. **Advanced: Reflection Agent** - Self-improvement pattern
+10. **Multi-Agent: Hierarchical Teams** - Complex orchestration
+
+---
+
+## Directory Structure
+
+```
+langgraph/
+├── README.md                    # Overview and getting started
+├── __init__.py
+├── basic/
+│   ├── README.md
+│   ├── hello_world/
+│   │   ├── README.md
+│   │   ├── run_worker.py
+│   │   └── run_workflow.py
+│   └── react_agent/
+│       ├── README.md
+│       ├── agent.py
+│       ├── tools.py
+│       ├── run_worker.py
+│       └── run_workflow.py
+├── human_in_loop/
+│   ├── README.md
+│   ├── approval_workflow/
+│   └── document_review/
+├── multi_agent/
+│   ├── README.md
+│   ├── supervisor/
+│   └── hierarchical_teams/
+├── rag/
+│   ├── README.md
+│   ├── agentic_rag/
+│   ├── corrective_rag/
+│   └── deep_research/
+├── customer_service/
+│   ├── README.md
+│   ├── support_agent/
+│   └── email_triage/
+├── planning/
+│   ├── README.md
+│   ├── plan_and_execute/
+│   └── travel_planner/
+├── code/
+│   ├── README.md
+│   └── code_assistant/
+└── advanced/
+    ├── README.md
+    ├── reflection/
+    ├── tree_of_thoughts/
+    └── long_running_workflow/
+```
+
+---
+
+## References
+
+### Official LangGraph Resources
+- [LangGraph Documentation](https://docs.langchain.com/oss/python/langgraph/overview)
+- [LangGraph GitHub Repository](https://github.com/langchain-ai/langgraph)
+- [LangGraph Tutorials](https://github.com/langchain-ai/langgraph/tree/main/docs/docs/tutorials)
+- [LangChain Academy - Intro to LangGraph](https://academy.langchain.com/courses/intro-to-langgraph)
+
+### Multi-Agent Resources
+- [LangGraph Multi-Agent Concepts](https://github.com/langchain-ai/langgraph/blob/main/docs/docs/concepts/multi_agent.md)
+- [LangGraph Supervisor Library](https://github.com/langchain-ai/langgraph-supervisor-py)
+- [Hierarchical Agent Teams Tutorial](https://langchain-ai.github.io/langgraph/tutorials/multi_agent/hierarchical_agent_teams/)
+
+### RAG and Research
+- [Agentic RAG Documentation](https://docs.langchain.com/oss/python/langgraph/agentic-rag)
+- [RAG Research Agent Template](https://github.com/langchain-ai/rag-research-agent-template)
+- [Deep Research Agent Article](https://towardsdatascience.com/langgraph-101-lets-build-a-deep-research-agent/)
+
+### Human-in-the-Loop
+- [Human-in-the-Loop How-To](https://langchain-ai.github.io/langgraph/how-tos/human_in_the_loop/)
+- [Human Feedback Discussion](https://github.com/langchain-ai/langgraph/discussions/1137)
+
+### Courses and Tutorials
+- [DeepLearning.AI - AI Agents in LangGraph](https://www.deeplearning.ai/short-courses/ai-agents-in-langgraph/)
+- [Real Python - LangGraph Tutorial](https://realpython.com/langgraph-python/)

langgraph_samples/__init__.py

@@ -0,0 +1 @@
+# Temporal LangGraph Samples

langgraph_samples/basic/README.md

@@ -0,0 +1,18 @@
+# Basic LangGraph Examples
+
+Simple examples to get started with LangGraph integrated with Temporal workflows.
+
+Before running these examples, be sure to review the [prerequisites and background on the integration](../README.md).
+
+## Examples
+
+### Hello World Agent
+
+Minimal LangGraph agent with Temporal integration. A single-node graph that processes a query.
+
+**Demonstrates:**
+- Basic plugin setup and graph registration
+- Simple workflow invocation
+- Activity-based node execution
+
+See [hello_world/README.md](./hello_world/README.md) for details.

langgraph_samples/basic/__init__.py

@@ -0,0 +1 @@
+# Basic LangGraph Examples