feat: A2A Real Estate Multi-Agent AgentCore example + README update

README.md

@@ -51,7 +51,28 @@ pip install -r requirements.txt
 
 ---
 
-#### [Amazon Bedrock AgentCore — Cost Optimization Agent](examples/amazon-bedrock-agentcore/) `NEW`
+#### [A2A Real Estate Multi-Agent — AgentCore Edition](examples/a2a-realestate-agentcore/) `NEW`
+*March 2026*
+
+A BMasterAI adaptation of the [AWS Labs A2A Real Estate sample](https://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/02-use-cases/A2A-realestate-agentcore-multiagents). Three Strands agents — Property Search, Property Booking, and a Coordinator — communicate over the A2A (Agent-to-Agent) protocol, with every tool call and A2A hop instrumented via BMasterAI structured telemetry.
+
+**Stack:** AWS Bedrock AgentCore, Strands, A2A Protocol, OAuth 2.0 (Cognito), BMasterAI
+
+**What it demonstrates:**
+- Multi-agent orchestration with the A2A protocol: coordinator delegates to specialized sub-agents at runtime
+- BMasterAI telemetry replacing custom loggers: `TOOL_USE`, `TASK_COMPLETE`, `TASK_ERROR` on every operation across all three agents
+- Bearer token forwarding from AgentCore Runtime context to sub-agent A2A calls
+- Local REPL mode for development + `BedrockAgentCoreApp` path for serverless deployment
+
+```bash
+# Start all three agents locally + interactive REPL
+pip install -r realestate_coordinator/requirements.txt
+python run_local.py
+```
+
+---
+
+#### [Amazon Bedrock AgentCore — Cost Optimization Agent](examples/amazon-bedrock-agentcore/)
 *February 2026*
 
 A Strands agent that monitors AWS spend, detects anomalies, forecasts costs, and analyzes service-level breakdowns — with BMasterAI structured telemetry logged on every agent action. Inspired by the awslabs/amazon-bedrock-agentcore-samples reference implementation.

examples/a2a-realestate-agentcore/README.md

@@ -0,0 +1,109 @@
+# A2A Real Estate Multi-Agent — BMasterAI Edition
+
+A **BMasterAI** adaptation of the AWS Labs
+[A2A Real Estate AgentCore sample](https://github.com/awslabs/amazon-bedrock-agentcore-samples/tree/main/02-use-cases/A2A-realestate-agentcore-multiagents).
+
+## Architecture
+
+```
+User / AgentCore Runtime
+        │
+        ▼
+Real Estate Coordinator (realestate_coordinator/agent.py)
+   ├─ search_properties ──► Property Search Agent  (port 5002)
+   │                              └─ search_properties tool
+   │                              └─ get_property_details tool
+   └─ book_property ──────► Property Booking Agent (port 5001)
+   └─ check_booking ──────►        └─ create_booking tool
+                                   └─ check_booking_status tool
+                                   └─ cancel_booking tool
+                                   └─ list_customer_bookings tool
+```
+
+All three agents use:
+- **[Strands Agents](https://github.com/strands-agents/sdk-python)** for tool-calling and LLM
+- **A2A protocol** for inter-agent communication
+- **BMasterAI** for structured telemetry (JSON logs + console)
+
+## Key Differences from the AWS Sample
+
+| AWS Sample | This Example |
+|---|---|
+| `common.utils.logging_config` (custom logger) | `bmasterai.logging` (structured telemetry) |
+| `setup_logging()` + `log_tool_execution()` + `log_error()` | `bm.log_event(EventType.TOOL_USE / TASK_COMPLETE / TASK_ERROR)` |
+| Separate `common/` module shared via `sys.path` | Each agent is self-contained |
+| AgentCore deployment scripts included | Deployment left to `bedrock-agentcore-starter-toolkit` |
+
+## Quick Start (Local)
+
+```bash
+# Install deps for all three agents
+pip install -r propertysearchagent/requirements.txt
+pip install -r propertybookingagent/requirements.txt
+pip install -r realestate_coordinator/requirements.txt
+
+# Start everything (sub-agents + coordinator REPL)
+python run_local.py
+```
+
+Or start agents individually in separate terminals:
+
+```bash
+# Terminal 1 — Booking Agent
+AGENT_PORT=5001 python propertybookingagent/agent.py
+
+# Terminal 2 — Search Agent
+AGENT_PORT=5002 python propertysearchagent/agent.py
+
+# Terminal 3 — Coordinator (interactive REPL)
+PROPERTY_SEARCH_AGENT_URL=http://localhost:5002 \
+PROPERTY_BOOKING_AGENT_URL=http://localhost:5001 \
+python realestate_coordinator/agent.py
+```
+
+## Example Queries
+
+```
+Find me a 2-bedroom apartment in New York under $4000/month
+
+Book PROP001 for Jane Doe, email jane@example.com, phone 555-1234, move-in 2025-09-01
+
+Check the status of booking BOOK-ABC12345
+
+List all bookings for jane@example.com
+
+Cancel booking BOOK-ABC12345 because I found a better option
+```
+
+## AgentCore Deployment
+
+Each agent directory contains everything needed to deploy to Amazon Bedrock AgentCore:
+
+1. **Sub-agents** (`propertysearchagent/`, `propertybookingagent/`) — deploy as A2A servers
+2. **Coordinator** — deploy with `--agentcore` flag or use `create_agentcore_app()` directly
+
+Set the following environment variables on the coordinator:
+- `PROPERTY_SEARCH_AGENT_URL` — AgentCore runtime URL of the search agent
+- `PROPERTY_BOOKING_AGENT_URL` — AgentCore runtime URL of the booking agent
+- `MODEL_ID` — Bedrock model ID (default: `us.anthropic.claude-3-5-haiku-20241022-v1:0`)
+
+The coordinator automatically forwards the incoming `Authorization` header to sub-agents
+when running inside AgentCore Runtime (`BedrockAgentCoreContext`).
+
+## BMasterAI Telemetry
+
+Every tool call emits structured events:
+
+```python
+bm.log_event(
+    agent_id=AGENT_ID,
+    event_type=EventType.TOOL_USE,        # or TASK_COMPLETE / TASK_ERROR
+    message="Tool: search_properties",
+    metadata={"location": "New York", "max_price": 4000},
+    duration_ms=45.2,
+)
+```
+
+Logs are written to:
+- Console (human-readable)
+- `logs/bmasterai.jsonl` (structured JSON for downstream processing)