Kiro Power - Amazon Bedrock AgentCore を学ぶ、もう一つの方法 (メモリー編)

Transcript

3. uv run agentcore memory create MyFirstAgentMemory --description "MyFirstAgent用のメモリリソース" --region us-east-1 --event- expiry-days 30 –wait ⠋ Resolving dependencies... Building myfirstagent @ file:///Users/y⠧ Preparing packages... (0/1) ⠧ Preparing packages... (0/1) ⠧ (1/1) Uninstalled 1 package in 4ms ░░░░░░░░░░░░░░░░░░░░ [0/1] Installing wheel ████████████████████ [1/1] myfirstagent==0. Installed 1 package in 12ms MemoryManager initialized for region: us-east-1 Creating memory: MyFirstAgentMemory... Created memory: MyFirstAgentMemory-NmqtwS9D5c Created memory MyFirstAgentMemory-NmqtwS9D5c, waiting for ACTIVE status... Waiting for memory MyFirstAgentMemory-NmqtwS9D5c to return to ACTIVE state and strategies to reach terminal states...
4. [06:47:53] Memory: manager.py:1029 CREATING (10s elapsed) [06:50:20] Memory: manager.py:1029 ACTIVE (157s elapsed) Memory MyFirstAgentMemory-NmqtwS9D5c is ACTIVE and all strategies are in terminal states (took 157 seconds) Memory is manager.py:1043 ACTIVE (took 157s) ObservabilityDeliveryManager initialized for region: us-east-1, account: 123456789012 Created log group: /aws/vendedlogs/bedrock-agentcore/memory/APPLICATION_LOGS/MyFirstAgentMemory-NmqtwS9D5c Logs delivery enabled for memory/MyFirstAgentMemory-NmqtwS9D5c Traces delivery enabled for memory/MyFirstAgentMemory-NmqtwS9D5c Observability enabled for memory/MyFirstAgentMemory-NmqtwS9D5c - logs: True, traces: True Observability enabled for memory MyFirstAgentMemory-NmqtwS9D5c Log group: /aws/vendedlogs/bedrock-agentcore/memory/APPLICATION_LOGS/MyFirstAgentMemory-NmqtwS9D5c ✓ Memory created successfully! Memory ID: MyFirstAgentMemory-NmqtwS9D5c Status: ACTIVE Region: us-east-1
8. uv run agentcore invoke --dev '{"prompt": "私の名前は田中太郎です。趣味はプログラミングで、好きな言 語はPythonです。よろしくお願いします。", "session_id": "memory_test_001", "actor_id": "tanaka_taro"}'
9. uv run agentcore invoke --dev '{"prompt": "私の名前と趣味を覚えていますか？", "session_id": "memory_test_001", "actor_id": "tanaka_taro"}'
10. uv run agentcore invoke --dev '{"prompt": "実は、私は機械学習エンジニアとして働いています。特にNLPが 専門分野です。", "session_id": "memory_test_001", "actor_id": "tanaka_taro"}'
11. uv run agentcore invoke --dev '{"prompt": "私のことを覚えていますか？", "session_id": "memory_test_002", "actor_id": "tanaka_taro"}'
12. uv run agentcore invoke --dev '{"prompt": "私の職業と専門分野について、詳し く教えてください", "session_id": "memory_test_001", "actor_id": "tanaka_taro"}'