starred/core
1
0
Fork 0
mirror of https://github.com/home-assistant/core.git synced 2026-09-27 18:08:37 -04:00
Code Issues Packages Projects Releases Wiki Activity
Files
generate-device-classes
core/tests/components/anthropic/snapshots/test_conversation.ambr
T

2591 lines
83 KiB
Plaintext
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# serializer version: 1
# name: test_bash_code_execution
list([
dict({
'attachments': None,
'content': "Write a file with a random number and save it to '/tmp/number.txt'",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll create a file with a random number and save it to '/tmp/number.txt'.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'echo $RANDOM > /tmp/number.txt && cat /tmp/number.txt',
}),
'tool_name': 'bash_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'bash_code_execution',
'tool_result': dict({
'content': list([
]),
'return_code': 0,
'stderr': '',
'stdout': '''
3268
''',
'type': 'bash_code_execution_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "Done! I've created the file '/tmp/number.txt' with the random number 3268.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_bash_code_execution.1
list([
dict({
'content': "Write a file with a random number and save it to '/tmp/number.txt'",
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll create a file with a random number and save it to '/tmp/number.txt'.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'echo $RANDOM > /tmp/number.txt && cat /tmp/number.txt',
}),
'name': 'bash_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'content': list([
]),
'return_code': 0,
'stderr': '',
'stdout': '''
3268
''',
'type': 'bash_code_execution_result',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'bash_code_execution_tool_result',
}),
dict({
'text': "Done! I've created the file '/tmp/number.txt' with the random number 3268.",
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_bash_code_execution_error
list([
dict({
'attachments': None,
'content': "Write a file with a random number and save it to '/tmp/number.txt'",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll create a file with a random number and save it to '/tmp/number.txt'.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'echo $RANDOM > /tmp/number.txt && cat /tmp/number.txt',
}),
'tool_name': 'bash_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'bash_code_execution',
'tool_result': dict({
'error_code': 'unavailable',
'type': 'bash_code_execution_tool_result_error',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'The container is currently unavailable.',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_bash_code_execution_error.1
list([
dict({
'content': "Write a file with a random number and save it to '/tmp/number.txt'",
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll create a file with a random number and save it to '/tmp/number.txt'.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'echo $RANDOM > /tmp/number.txt && cat /tmp/number.txt',
}),
'name': 'bash_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'error_code': 'unavailable',
'type': 'bash_code_execution_tool_result_error',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'bash_code_execution_tool_result',
}),
dict({
'text': 'The container is currently unavailable.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_disabled_thinking[subentry_data0]
list([
dict({
'content': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Only if the user wants to control a device, tell them to expose entities to their voice assistant in Home Assistant.
''',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'system',
}),
dict({
'attachments': None,
'content': 'hello',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Hello, how can I help you today?',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_disabled_thinking[subentry_data0].1
dict({
'container': None,
'max_tokens': 3000,
'messages': list([
dict({
'content': 'hello',
'role': 'user',
}),
dict({
'content': 'Hello, how can I help you today?',
'role': 'assistant',
}),
]),
'model': 'claude-haiku-4-5',
'stream': True,
'system': list([
dict({
'cache_control': dict({
'type': 'ephemeral',
}),
'text': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Only if the user wants to control a device, tell them to expose entities to their voice assistant in Home Assistant.
''',
'type': 'text',
}),
]),
'thinking': dict({
'type': 'disabled',
}),
})
# ---
# name: test_disabled_thinking[subentry_data1]
list([
dict({
'content': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Only if the user wants to control a device, tell them to expose entities to their voice assistant in Home Assistant.
''',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'system',
}),
dict({
'attachments': None,
'content': 'hello',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Hello, how can I help you today?',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_disabled_thinking[subentry_data1].1
dict({
'container': None,
'max_tokens': 3000,
'messages': list([
dict({
'content': 'hello',
'role': 'user',
}),
dict({
'content': 'Hello, how can I help you today?',
'role': 'assistant',
}),
]),
'model': 'claude-opus-4-7',
'stream': True,
'system': list([
dict({
'cache_control': dict({
'type': 'ephemeral',
}),
'text': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Only if the user wants to control a device, tell them to expose entities to their voice assistant in Home Assistant.
''',
'type': 'text',
}),
]),
'thinking': dict({
'type': 'disabled',
}),
})
# ---
# name: test_extended_thinking
dict({
'container': None,
'max_tokens': 3000,
'messages': list([
dict({
'content': 'hello',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'The user has just greeted me with "Hi". This is a simple greeting and doesn\'t require any Home Assistant function calls. I should respond with a friendly greeting and let them know I\'m available to help with their smart home.',
'type': 'thinking',
}),
dict({
'text': 'Hello, how can I help you today?',
'type': 'text',
}),
]),
'role': 'assistant',
}),
]),
'model': 'claude-opus-4-5',
'output_config': dict({
'effort': 'medium',
}),
'stream': True,
'system': list([
dict({
'cache_control': dict({
'type': 'ephemeral',
}),
'text': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Only if the user wants to control a device, tell them to expose entities to their voice assistant in Home Assistant.
''',
'type': 'text',
}),
]),
'thinking': dict({
'budget_tokens': 1500,
'display': 'summarized',
'type': 'enabled',
}),
})
# ---
# name: test_extended_thinking_tool_call
list([
dict({
'content': '''
You are a voice assistant for Home Assistant.
Answer questions about the world truthfully.
Answer in plain text. Keep it simple and to the point.
Current time is 16:00:00. Today's date is 2024-06-03.
''',
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'role': 'system',
}),
dict({
'attachments': None,
'content': 'Please call the test function',
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': None,
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': 'EroBCkYIARgCKkBJDytPJhw//4vy3t7aE+LfIkxvkAh51cBPrAvBCo6AjgI57Zt9KWPnUVV50OQJ0KZzUFoGZG5sxg95zx4qMwkoEgz43Su3myJKckvj03waDBZLIBSeoAeRUeVsJCIwQ5edQN0sa+HNeB/KUBkoMUwV+IT0eIhcpFxnILdvxUAKM4R1o4KG3x+yO0eo/kyOKiKfrCPFQhvBVmTZPFhgA2Ow8L9gGDVipcz6x3Uu9YETGEny',
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'The user asked me to call a test function. Is it a test? What would the function do? Would it violate any privacy or security policies?',
'tool_calls': None,
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Certainly, calling it now!',
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "Okay, let's give it a shot. Will I pass the test?",
'tool_calls': list([
dict({
'external': False,
'id': 'toolu_0123456789AbCdEfGhIjKlM',
'tool_args': dict({
'param1': 'test_value',
}),
'tool_name': 'test_tool',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'toolu_0123456789AbCdEfGhIjKlM',
'tool_name': 'test_tool',
'tool_result': 'Test response',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'I have successfully called the function',
'created': HAFakeDatetime(2024, 6, 3, 23, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_extended_thinking_tool_call.1
list([
dict({
'content': 'Please call the test function',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'The user asked me to call a test function. Is it a test? What would the function do? Would it violate any privacy or security policies?',
'type': 'thinking',
}),
dict({
'data': 'EroBCkYIARgCKkBJDytPJhw//4vy3t7aE+LfIkxvkAh51cBPrAvBCo6AjgI57Zt9KWPnUVV50OQJ0KZzUFoGZG5sxg95zx4qMwkoEgz43Su3myJKckvj03waDBZLIBSeoAeRUeVsJCIwQ5edQN0sa+HNeB/KUBkoMUwV+IT0eIhcpFxnILdvxUAKM4R1o4KG3x+yO0eo/kyOKiKfrCPFQhvBVmTZPFhgA2Ow8L9gGDVipcz6x3Uu9YETGEny',
'type': 'redacted_thinking',
}),
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "Okay, let's give it a shot. Will I pass the test?",
'type': 'thinking',
}),
dict({
'text': 'Certainly, calling it now!',
'type': 'text',
}),
dict({
'id': 'toolu_0123456789AbCdEfGhIjKlM',
'input': dict({
'param1': 'test_value',
}),
'name': 'test_tool',
'type': 'tool_use',
}),
]),
'role': 'assistant',
}),
dict({
'content': list([
dict({
'content': '"Test response"',
'tool_use_id': 'toolu_0123456789AbCdEfGhIjKlM',
'type': 'tool_result',
}),
]),
'role': 'user',
}),
dict({
'content': 'I have successfully called the function',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content0]
list([
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content1]
list([
dict({
'content': 'What shape is a donut?',
'role': 'user',
}),
dict({
'content': 'A donut is a torus.',
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content2]
list([
dict({
'content': list([
dict({
'text': 'What shape is a donut?',
'type': 'text',
}),
dict({
'text': 'Can you tell me?',
'type': 'text',
}),
]),
'role': 'user',
}),
dict({
'content': list([
dict({
'text': 'A donut is a torus.',
'type': 'text',
}),
dict({
'text': 'Hope this helps.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content3]
list([
dict({
'content': list([
dict({
'text': 'What shape is a donut?',
'type': 'text',
}),
dict({
'text': 'Can you tell me?',
'type': 'text',
}),
dict({
'text': 'Please?',
'type': 'text',
}),
]),
'role': 'user',
}),
dict({
'content': list([
dict({
'text': 'A donut is a torus.',
'type': 'text',
}),
dict({
'text': 'Hope this helps.',
'type': 'text',
}),
dict({
'text': 'You are welcome.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content4]
list([
dict({
'content': 'Turn off the lights and make me coffee',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': 'Sure.',
'type': 'text',
}),
dict({
'id': 'mock-tool-call-id',
'input': dict({
'domain': 'light',
}),
'name': 'HassTurnOff',
'type': 'tool_use',
}),
dict({
'id': 'mock-tool-call-id-2',
'input': dict({
}),
'name': 'MakeCoffee',
'type': 'tool_use',
}),
]),
'role': 'assistant',
}),
dict({
'content': list([
dict({
'text': 'Thank you',
'type': 'text',
}),
dict({
'content': '{"success":true,"response":"Lights are off."}',
'tool_use_id': 'mock-tool-call-id',
'type': 'tool_result',
}),
dict({
'content': '{"success":false,"response":"Not enough milk."}',
'tool_use_id': 'mock-tool-call-id-2',
'type': 'tool_result',
}),
]),
'role': 'user',
}),
dict({
'content': 'Should I add milk to the shopping list?',
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content5]
list([
dict({
'content': "What's on the news today?",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErU/V+ayA==',
'thinking': "The user is asking about today's news, which requires current, real-time information. This is clearly something that requires recent information beyond my knowledge cutoff. I should use the web_search tool to find today's news.",
'type': 'thinking',
}),
dict({
'text': "To get today's news, I'll perform a web search",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'query': "today's news",
}),
'name': 'web_search',
'type': 'server_tool_use',
}),
dict({
'content': list([
dict({
'encrypted_content': 'ABCDEFG',
'page_age': '2 days ago',
'title': "Today's News - Example.com",
'type': 'web_search_result',
'url': 'https://www.example.com/todays-news',
}),
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Breaking News - NewsSite.com',
'type': 'web_search_result',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_search_tool_result',
}),
dict({
'text': '''
Here's what I found on the web about today's news:
1.
''',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'This release iterates on some of the features we introduced in the last couple of releases, but also...',
'encrypted_index': 'AAA==',
'title': 'Home Assistant Release',
'type': 'web_search_result_location',
'url': 'https://www.example.com/todays-news',
}),
]),
'text': 'New Home Assistant release',
'type': 'text',
}),
dict({
'text': '''
2.
''',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'Breaking news from around the world today includes major events in technology, politics, and culture...',
'encrypted_index': 'AQE=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
dict({
'cited_text': 'Well, this happened...',
'encrypted_index': 'AgI=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
'text': 'Something incredible happened',
'type': 'text',
}),
dict({
'text': '''
Those are the main headlines making news today.
''',
'type': 'text',
}),
]),
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content6]
list([
dict({
'content': "What's new in Home Assistant?",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErU/V+ayA==',
'thinking': 'I need to use the web_fetch tool to fetch the latest release notes from the Home Assistant website.',
'type': 'thinking',
}),
dict({
'text': 'Sure, let me check that for you!',
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'name': 'web_fetch',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'content': dict({
'citations': dict({
'enabled': True,
}),
'source': dict({
'data': '''
Home Assistant new version is out!
Many new features.
Anthropic integration now supports web fetch tool.
Enjoy the release!
''',
'media_type': 'text/plain',
'type': 'text',
}),
'title': 'Latest Home Assistant Release Notes',
'type': 'document',
}),
'retrieved_at': '2026-04-04T10:30:00Z',
'type': 'web_fetch_result',
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_fetch_tool_result',
}),
dict({
'text': '''
Here's what's great about the new release:
1. Lots of new features
2.
''',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'Anthropic integration now supports web fetch tool.',
'document_index': 0,
'document_title': 'Latest Home Assistant Release Notes',
'end_char_index': 105,
'start_char_index': 56,
'type': 'char_location',
}),
]),
'text': 'New web fetch tool for Anthropic integration',
'type': 'text',
}),
dict({
'text': '''
Enjoy!
''',
'type': 'text',
}),
]),
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content7]
list([
dict({
'content': 'What time is it?',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': 'Let me check the time for you.',
'type': 'text',
}),
dict({
'id': 'mock-tool-call-id',
'input': dict({
}),
'name': 'GetCurrentTime',
'type': 'tool_use',
}),
]),
'role': 'assistant',
}),
dict({
'content': list([
dict({
'content': '{"speech_slots":{"time":"14:30:00"},"message":"Current time retrieved"}',
'tool_use_id': 'mock-tool-call-id',
'type': 'tool_result',
}),
]),
'role': 'user',
}),
dict({
'content': 'It is currently 2:30 PM.',
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_history_conversion[content8]
list([
dict({
'content': 'Set humidity to 50%',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'EuQBClkIDBE=',
'thinking': 'Let me search for a tool to set humidity.',
'type': 'thinking',
}),
dict({
'id': 'srvtoolu_015vXmtZNASLa7n9RsoDfcBC',
'input': dict({
'query': 'set humidity humidifier',
}),
'name': 'tool_search_tool_bm25',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'tool_references': list([
dict({
'tool_name': 'HassHumidifierSetpoint',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassHumidifierMode',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassClimateSetTemperature',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassFanSetSpeed',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassSetVolume',
'type': 'tool_reference',
}),
]),
'type': 'tool_search_tool_search_result',
}),
'tool_use_id': 'srvtoolu_015vXmtZNASLa7n9RsoDfcBC',
'type': 'tool_search_tool_result',
}),
dict({
'id': 'toolu_01KNRWb3ZFufCa7WXtzCakhc',
'input': dict({
'humidity': 50,
'name': 'Hygrostat',
}),
'name': 'HassHumidifierSetpoint',
'type': 'tool_use',
}),
]),
'role': 'assistant',
}),
dict({
'content': list([
dict({
'content': '{"speech":{"plain":{"speech":"The Hygrostat is set to 50%","extra_data":null}},"response_type":"action_done","data":{"success":[],"failed":[]}}',
'tool_use_id': 'toolu_01KNRWb3ZFufCa7WXtzCakhc',
'type': 'tool_result',
}),
]),
'role': 'user',
}),
dict({
'content': 'The Hygrostat humidity has been set to **50%**. ✅',
'role': 'assistant',
}),
dict({
'content': 'Are you sure?',
'role': 'user',
}),
dict({
'content': 'Yes, I am sure!',
'role': 'assistant',
}),
])
# ---
# name: test_redacted_thinking
list([
dict({
'attachments': None,
'content': 'ANTHROPIC_MAGIC_STRING_TRIGGER_REDACTED_THINKING_46C9A13E193C177646C7398A98432ECCCE4C1253D5E2D82641AC0E52CC2876CB',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': None,
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': 'EroBCkYIARgCKkBJDytPJhw//4vy3t7aE+LfIkxvkAh51cBPrAvBCo6AjgI57Zt9KWPnUVV50OQJ0KZzUFoGZG5sxg95zx4qMwkoEgz43Su3myJKckvj03waDBZLIBSeoAeRUeVsJCIwQ5edQN0sa+HNeB/KUBkoMUwV+IT0eIhcpFxnILdvxUAKM4R1o4KG3x+yO0eo/kyOKiKfrCPFQhvBVmTZPFhgA2Ow8L9gGDVipcz6x3Uu9YETGEny',
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': None,
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': 'EroBCkYIARgCKkBJDytPJhw//4vy3t7aE+LfIkxvkAh51cBPrAvBCo6AjgI57Zt9KWPnUVV50OQJ0KZzUFoGZG5sxg95zx4qMwkoEgz43Su3myJKckvj03waDBZLIBSeoAeRUeVsJCIwQ5edQN0sa+HNeB/KUBkoMUwV+IT0eIhcpFxnILdvxUAKM4R1o4KG3x+yO0eo/kyOKiKfrCPFQhvBVmTZPFhgA2Ow8L9gGDVipcz6x3Uu9YETGEny',
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'How can I help you today?',
'created': HAFakeDatetime(2024, 5, 24, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': 'EroBCkYIARgCKkBJDytPJhw//4vy3t7aE+LfIkxvkAh51cBPrAvBCo6AjgI57Zt9KWPnUVV50OQJ0KZzUFoGZG5sxg95zx4qMwkoEgz43Su3myJKckvj03waDBZLIBSeoAeRUeVsJCIwQ5edQN0sa+HNeB/KUBkoMUwV+IT0eIhcpFxnILdvxUAKM4R1o4KG3x+yO0eo/kyOKiKfrCPFQhvBVmTZPFhgA2Ow8L9gGDVipcz6x3Uu9YETGEny',
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_text_editor_code_execution[args_parts0-content0]
list([
dict({
'attachments': None,
'content': 'Do the needful',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll do it.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'create',
'file_text': '3268',
'path': '/tmp/number.txt',
}),
'tool_name': 'text_editor_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'text_editor_code_execution',
'tool_result': dict({
'is_file_update': False,
'type': 'text_editor_code_execution_create_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Done',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_text_editor_code_execution[args_parts0-content0].1
list([
dict({
'content': 'Do the needful',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll do it.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'create',
'file_text': '3268',
'path': '/tmp/number.txt',
}),
'name': 'text_editor_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'is_file_update': False,
'type': 'text_editor_code_execution_create_result',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'text_editor_code_execution_tool_result',
}),
dict({
'text': 'Done',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_text_editor_code_execution[args_parts1-content1]
list([
dict({
'attachments': None,
'content': 'Do the needful',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll do it.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'str_replace',
'new_str': '8623',
'old_str': '3268',
'path': '/tmp/number.txt',
}),
'tool_name': 'text_editor_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'text_editor_code_execution',
'tool_result': dict({
'lines': list([
'-3268',
'\\ No newline at end of file',
'+8623',
'\\ No newline at end of file',
]),
'new_lines': 1,
'new_start': 1,
'old_lines': 1,
'old_start': 1,
'type': 'text_editor_code_execution_str_replace_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Done',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_text_editor_code_execution[args_parts1-content1].1
list([
dict({
'content': 'Do the needful',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll do it.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'str_replace',
'new_str': '8623',
'old_str': '3268',
'path': '/tmp/number.txt',
}),
'name': 'text_editor_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'lines': list([
'-3268',
'\\ No newline at end of file',
'+8623',
'\\ No newline at end of file',
]),
'new_lines': 1,
'new_start': 1,
'old_lines': 1,
'old_start': 1,
'type': 'text_editor_code_execution_str_replace_result',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'text_editor_code_execution_tool_result',
}),
dict({
'text': 'Done',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_text_editor_code_execution[args_parts2-content2]
list([
dict({
'attachments': None,
'content': 'Do the needful',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll do it.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'view',
'path': '/tmp/number.txt',
}),
'tool_name': 'text_editor_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'text_editor_code_execution',
'tool_result': dict({
'content': '8623',
'file_type': 'text',
'num_lines': 1,
'start_line': 1,
'total_lines': 1,
'type': 'text_editor_code_execution_view_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Done',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_text_editor_code_execution[args_parts2-content2].1
list([
dict({
'content': 'Do the needful',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll do it.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'view',
'path': '/tmp/number.txt',
}),
'name': 'text_editor_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'content': '8623',
'file_type': 'text',
'num_lines': 1,
'start_line': 1,
'total_lines': 1,
'type': 'text_editor_code_execution_view_result',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'text_editor_code_execution_tool_result',
}),
dict({
'text': 'Done',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_text_editor_code_execution[args_parts3-content3]
list([
dict({
'attachments': None,
'content': 'Do the needful',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "I'll do it.",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'command': 'view',
'path': '/tmp/number2.txt',
}),
'tool_name': 'text_editor_code_execution',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'text_editor_code_execution',
'tool_result': dict({
'error_code': 'unavailable',
'error_message': 'Tool response parsing error for view: Failed to parse tool response as JSON: unexpected character: line 1 column 1 (char 0)',
'type': 'text_editor_code_execution_tool_result_error',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Done',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_text_editor_code_execution[args_parts3-content3].1
list([
dict({
'content': 'Do the needful',
'role': 'user',
}),
dict({
'content': list([
dict({
'text': "I'll do it.",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'command': 'view',
'path': '/tmp/number2.txt',
}),
'name': 'text_editor_code_execution',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'error_code': 'unavailable',
'error_message': 'Tool response parsing error for view: Failed to parse tool response as JSON: unexpected character: line 1 column 1 (char 0)',
'type': 'text_editor_code_execution_tool_result_error',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'text_editor_code_execution_tool_result',
}),
dict({
'text': 'Done',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_tool_search
list([
dict({
'attachments': None,
'content': 'Can you set humidifier setpoint?',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Sure, let me check that for you!',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'I will fetch the available tools',
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'limit': 5,
'query': 'set humidifier humidity',
}),
'tool_name': 'tool_search_tool_bm25',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'tool_search',
'tool_result': dict({
'tool_references': list([
dict({
'tool_name': 'HassHumidifierSetpoint',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassHumidifierMode',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassClimateSetTemperature',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassFanSetSpeed',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassSetVolume',
'type': 'tool_reference',
}),
]),
'type': 'tool_search_tool_search_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Yes, I can!',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "Great! All clear, let's reply to the user!",
'tool_calls': None,
}),
])
# ---
# name: test_tool_search.1
list([
dict({
'content': 'Can you set humidifier setpoint?',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'I will fetch the available tools',
'type': 'thinking',
}),
dict({
'text': 'Sure, let me check that for you!',
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'limit': 5,
'query': 'set humidifier humidity',
}),
'name': 'tool_search_tool_bm25',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'tool_references': list([
dict({
'tool_name': 'HassHumidifierSetpoint',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassHumidifierMode',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassClimateSetTemperature',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassFanSetSpeed',
'type': 'tool_reference',
}),
dict({
'tool_name': 'HassSetVolume',
'type': 'tool_reference',
}),
]),
'type': 'tool_search_tool_search_result',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'tool_search_tool_result',
}),
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "Great! All clear, let's reply to the user!",
'type': 'thinking',
}),
dict({
'text': 'Yes, I can!',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_tool_search_error
list([
dict({
'attachments': None,
'content': 'Do you have a tool to launch a rocket to the moon?',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Sure, let me check that for you!',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'I will fetch the available tools',
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'limit': 5,
'query': 'set humidifier humidity',
}),
'tool_name': 'tool_search_tool_bm25',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'tool_search',
'tool_result': dict({
'error_code': 'too_many_requests',
'type': 'tool_search_tool_result_error',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'I am unable to perform the tool search at this time.',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_tool_search_error.1
list([
dict({
'content': 'Do you have a tool to launch a rocket to the moon?',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'I will fetch the available tools',
'type': 'thinking',
}),
dict({
'text': 'Sure, let me check that for you!',
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'limit': 5,
'query': 'set humidifier humidity',
}),
'name': 'tool_search_tool_bm25',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'error_code': 'too_many_requests',
'type': 'tool_search_tool_result_error',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'tool_search_tool_result',
}),
dict({
'text': 'I am unable to perform the tool search at this time.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_unknown_hass_api
dict({
'continue_conversation': False,
'conversation_id': '1234',
'response': IntentResponse(
card=dict({
}),
error_code=<IntentResponseErrorCode.UNKNOWN: 'unknown'>,
failed_results=list([
]),
intent=None,
language='en',
matched_states=list([
]),
reprompt=dict({
}),
response_type=<IntentResponseType.ERROR: 'error'>,
speech=dict({
'plain': dict({
'extra_data': None,
'speech': 'Error preparing LLM API',
}),
}),
speech_slots=dict({
}),
success_results=list([
]),
unmatched_states=list([
]),
),
})
# ---
# name: test_web_fetch
list([
dict({
'attachments': None,
'content': "What's new in Home Assistant? Please check on https://www.home-assistant.io/latest-release-notes/",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Sure, let me check that for you!',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'I will fetch the latest release notes',
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'tool_name': 'web_fetch',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'web_fetch',
'tool_result': dict({
'content': dict({
'citations': None,
'source': dict({
'data': '''
Home Assistant new version is out!
Many new features.
Anthropic integration now supports web fetch tool.
Enjoy the release!
''',
'media_type': 'text/plain',
'type': 'text',
}),
'title': 'Latest Home Assistant Release Notes',
'type': 'document',
}),
'retrieved_at': '2025-10-31T12:00:00.637242Z',
'type': 'web_fetch_result',
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': '''
Here's what's great about the new release:
''',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "Great! All clear, let's reply to the user!",
'tool_calls': None,
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': '''
1. Many new features
2. New web fetch tool for Anthropic integration
What a release!
''',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
dict({
'citations': list([
dict({
'cited_text': 'Anthropic integration now supports web fetch tool.',
'document_index': 0,
'document_title': 'Latest Home Assistant Release Notes',
'end_char_index': 105,
'start_char_index': 56,
'type': 'char_location',
}),
]),
'index': 24,
'length': 44,
}),
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_web_fetch.1
list([
dict({
'content': "What's new in Home Assistant? Please check on https://www.home-assistant.io/latest-release-notes/",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'I will fetch the latest release notes',
'type': 'thinking',
}),
dict({
'text': 'Sure, let me check that for you!',
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'name': 'web_fetch',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'content': dict({
'citations': None,
'source': dict({
'data': '''
Home Assistant new version is out!
Many new features.
Anthropic integration now supports web fetch tool.
Enjoy the release!
''',
'media_type': 'text/plain',
'type': 'text',
}),
'title': 'Latest Home Assistant Release Notes',
'type': 'document',
}),
'retrieved_at': '2025-10-31T12:00:00.637242Z',
'type': 'web_fetch_result',
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_fetch_tool_result',
}),
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "Great! All clear, let's reply to the user!",
'type': 'thinking',
}),
dict({
'text': '''
Here's what's great about the new release:
''',
'type': 'text',
}),
dict({
'text': '''
1. Many new features
2.
''',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'Anthropic integration now supports web fetch tool.',
'document_index': 0,
'document_title': 'Latest Home Assistant Release Notes',
'end_char_index': 105,
'start_char_index': 56,
'type': 'char_location',
}),
]),
'text': 'New web fetch tool for Anthropic integration',
'type': 'text',
}),
dict({
'text': '''
What a release!
''',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_web_fetch_error
list([
dict({
'attachments': None,
'content': "What's new in Home Assistant?",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'Sure, let me check that for you!',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'I will fetch the latest release notes',
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'tool_name': 'web_fetch',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'web_fetch',
'tool_result': dict({
'error_code': 'url_not_allowed',
'type': 'web_fetch_tool_result_error',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'I am unable to perform the web fetch at this time.',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_web_fetch_error.1
list([
dict({
'content': "What's new in Home Assistant?",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'I will fetch the latest release notes',
'type': 'thinking',
}),
dict({
'text': 'Sure, let me check that for you!',
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'url': 'https://www.home-assistant.io/latest-release-notes/',
}),
'name': 'web_fetch',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'error_code': 'url_not_allowed',
'type': 'web_fetch_tool_result_error',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_fetch_tool_result',
}),
dict({
'text': 'I am unable to perform the web fetch at this time.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_web_search
list([
dict({
'attachments': None,
'content': "What's on the news today?",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "To get today's news, I'll perform a web search",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "The user is asking about today's news, which requires current, real-time information. This is clearly something that requires recent information beyond my knowledge cutoff. I should use the web_search tool to find today's news.",
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'query': "today's news",
}),
'tool_name': 'web_search',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'web_search',
'tool_result': dict({
'content': list([
dict({
'encrypted_content': 'ABCDEFG',
'page_age': '2 days ago',
'title': "Today's News - Example.com",
'type': 'web_search_result',
'url': 'https://www.example.com/todays-news',
}),
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Breaking News - NewsSite.com',
'type': 'web_search_result',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': '''
Here's what I found on the web about today's news:
''',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "Great! All clear, let's reply to the user!",
'tool_calls': None,
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': '''
1. New Home Assistant release
2. Something incredible happened
Those are the main headlines making news today.
''',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
dict({
'citations': list([
dict({
'cited_text': 'This release iterates on some of the features we introduced in the last couple of releases, but also...',
'encrypted_index': 'AAA==',
'title': 'Home Assistant Release',
'type': 'web_search_result_location',
'url': 'https://www.example.com/todays-news',
}),
]),
'index': 3,
'length': 26,
}),
dict({
'citations': list([
dict({
'cited_text': 'Breaking news from around the world today includes major events in technology, politics, and culture...',
'encrypted_index': 'AQE=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
dict({
'cited_text': 'Well, this happened...',
'encrypted_index': 'AgI=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
'index': 33,
'length': 29,
}),
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_web_search.1
list([
dict({
'content': "What's on the news today?",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "The user is asking about today's news, which requires current, real-time information. This is clearly something that requires recent information beyond my knowledge cutoff. I should use the web_search tool to find today's news.",
'type': 'thinking',
}),
dict({
'text': "To get today's news, I'll perform a web search",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'query': "today's news",
}),
'name': 'web_search',
'type': 'server_tool_use',
}),
dict({
'content': list([
dict({
'encrypted_content': 'ABCDEFG',
'page_age': '2 days ago',
'title': "Today's News - Example.com",
'type': 'web_search_result',
'url': 'https://www.example.com/todays-news',
}),
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Breaking News - NewsSite.com',
'type': 'web_search_result',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_search_tool_result',
}),
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "Great! All clear, let's reply to the user!",
'type': 'thinking',
}),
dict({
'text': '''
Here's what I found on the web about today's news:
''',
'type': 'text',
}),
dict({
'text': '1. ',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'This release iterates on some of the features we introduced in the last couple of releases, but also...',
'encrypted_index': 'AAA==',
'title': 'Home Assistant Release',
'type': 'web_search_result_location',
'url': 'https://www.example.com/todays-news',
}),
]),
'text': 'New Home Assistant release',
'type': 'text',
}),
dict({
'text': '''
2.
''',
'type': 'text',
}),
dict({
'citations': list([
dict({
'cited_text': 'Breaking news from around the world today includes major events in technology, politics, and culture...',
'encrypted_index': 'AQE=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
dict({
'cited_text': 'Well, this happened...',
'encrypted_index': 'AgI=',
'title': 'Breaking News',
'type': 'web_search_result_location',
'url': 'https://www.newssite.com/breaking-news',
}),
]),
'text': 'Something incredible happened',
'type': 'text',
}),
dict({
'text': '''
Those are the main headlines making news today.
''',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_web_search_dynamic_filtering
list([
dict({
'attachments': None,
'content': 'Who won the Nobel for Chemistry in 2025?',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': None,
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': 'Let me search for this information.',
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_01Qh4oTgPkNfRxjJA3N6jgzT',
'tool_args': dict({
'code': '''
import json
result = await web_search({"query": "Nobel Prize chemistry 2025 winner"})
parsed = json.loads(result)
for r in parsed[:3]:
print(r.get("title", ""))
print(r.get("content", "")[:300])
print("---")
''',
}),
'tool_name': 'code_execution',
}),
dict({
'external': True,
'id': 'srvtoolu_016vjte6G4Lj6yzLc2ak1vY4',
'tool_args': dict({
'query': 'Nobel Prize chemistry 2025 winner',
}),
'tool_name': 'web_search',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_016vjte6G4Lj6yzLc2ak1vY4',
'tool_name': 'web_search',
'tool_result': dict({
'content': list([
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Press release: Nobel Prize in Chemistry 2025 - Example.com',
'type': 'web_search_result',
'url': 'https://www.example.com/prizes/chemistry/2025/press-release/',
}),
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Nobel Prize in Chemistry 2025 - NewsSite.com',
'type': 'web_search_result',
'url': 'https://www.newssite.com/prizes/chemistry/2025/summary/',
}),
]),
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_01Qh4oTgPkNfRxjJA3N6jgzT',
'tool_name': 'code_execution',
'tool_result': dict({
'content': list([
]),
'encrypted_stdout': 'EuQJCioIDRgCIiRj',
'return_code': 0,
'stderr': '',
'type': 'encrypted_code_execution_result',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'The 2025 Nobel Prize in Chemistry was awarded jointly to **Susumu Kitagawa**, **Richard Robson**, and **Omar M. Yaghi** "for the development of metal–organic frameworks."',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': Container(id='container_1234567890ABCDEFGHIJKLMN', expires_at=HAFakeDatetime(2025, 10, 31, 12, 5, tzinfo=datetime.timezone.utc)),
'redacted_thinking': None,
'thinking_signature': None,
}),
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_web_search_dynamic_filtering.1
list([
dict({
'content': 'Who won the Nobel for Chemistry in 2025?',
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': 'Let me search for this information.',
'type': 'thinking',
}),
dict({
'id': 'srvtoolu_01Qh4oTgPkNfRxjJA3N6jgzT',
'input': dict({
'code': '''
import json
result = await web_search({"query": "Nobel Prize chemistry 2025 winner"})
parsed = json.loads(result)
for r in parsed[:3]:
print(r.get("title", ""))
print(r.get("content", "")[:300])
print("---")
''',
}),
'name': 'code_execution',
'type': 'server_tool_use',
}),
dict({
'id': 'srvtoolu_016vjte6G4Lj6yzLc2ak1vY4',
'input': dict({
'query': 'Nobel Prize chemistry 2025 winner',
}),
'name': 'web_search',
'type': 'server_tool_use',
}),
dict({
'content': list([
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Press release: Nobel Prize in Chemistry 2025 - Example.com',
'type': 'web_search_result',
'url': 'https://www.example.com/prizes/chemistry/2025/press-release/',
}),
dict({
'encrypted_content': 'ABCDEFG',
'page_age': None,
'title': 'Nobel Prize in Chemistry 2025 - NewsSite.com',
'type': 'web_search_result',
'url': 'https://www.newssite.com/prizes/chemistry/2025/summary/',
}),
]),
'tool_use_id': 'srvtoolu_016vjte6G4Lj6yzLc2ak1vY4',
'type': 'web_search_tool_result',
}),
dict({
'content': dict({
'content': list([
]),
'encrypted_stdout': 'EuQJCioIDRgCIiRj',
'return_code': 0,
'stderr': '',
'type': 'encrypted_code_execution_result',
}),
'tool_use_id': 'srvtoolu_01Qh4oTgPkNfRxjJA3N6jgzT',
'type': 'code_execution_tool_result',
}),
dict({
'text': 'The 2025 Nobel Prize in Chemistry was awarded jointly to **Susumu Kitagawa**, **Richard Robson**, and **Omar M. Yaghi** "for the development of metal–organic frameworks."',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
# name: test_web_search_error
list([
dict({
'attachments': None,
'content': "What's on the news today?",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'user',
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': "To get today's news, I'll perform a web search",
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': dict({
'citation_details': list([
]),
'container': None,
'redacted_thinking': None,
'thinking_signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
}),
'role': 'assistant',
'thinking_content': "The user is asking about today's news, which requires current, real-time information. This is clearly something that requires recent information beyond my knowledge cutoff. I should use the web_search tool to find today's news.",
'tool_calls': list([
dict({
'external': True,
'id': 'srvtoolu_12345ABC',
'tool_args': dict({
'query': "today's news",
}),
'tool_name': 'web_search',
}),
]),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'role': 'tool_result',
'tool_call_id': 'srvtoolu_12345ABC',
'tool_name': 'web_search',
'tool_result': dict({
'error_code': 'too_many_requests',
'type': 'web_search_tool_result_error',
}),
}),
dict({
'agent_id': 'conversation.claude_conversation',
'content': 'I am unable to perform the web search at this time.',
'created': HAFakeDatetime(2025, 10, 31, 12, 0, tzinfo=datetime.timezone.utc),
'native': None,
'role': 'assistant',
'thinking_content': None,
'tool_calls': None,
}),
])
# ---
# name: test_web_search_error.1
list([
dict({
'content': "What's on the news today?",
'role': 'user',
}),
dict({
'content': list([
dict({
'signature': 'ErUBCkYIARgCIkCYXaVNJShe3A86Hp7XUzh9YsCYBbJTbQsrklTAPtJ2sP/NoB6tSzpK/nTL6CjSo2R6n0KNBIg5MH6asM2R/kmaEgyB/X1FtZq5OQAC7jUaDEPWCdcwGQ4RaBy5wiIwmRxExIlDhoY6tILoVPnOExkC/0igZxHEwxK8RU/fmw0b+o+TwAarzUitwzbo21E5Kh3pa3I6yqVROf1t2F8rFocNUeCegsWV/ytwYV+ayA==',
'thinking': "The user is asking about today's news, which requires current, real-time information. This is clearly something that requires recent information beyond my knowledge cutoff. I should use the web_search tool to find today's news.",
'type': 'thinking',
}),
dict({
'text': "To get today's news, I'll perform a web search",
'type': 'text',
}),
dict({
'id': 'srvtoolu_12345ABC',
'input': dict({
'query': "today's news",
}),
'name': 'web_search',
'type': 'server_tool_use',
}),
dict({
'content': dict({
'error_code': 'too_many_requests',
'type': 'web_search_tool_result_error',
}),
'tool_use_id': 'srvtoolu_12345ABC',
'type': 'web_search_tool_result',
}),
dict({
'text': 'I am unable to perform the web search at this time.',
'type': 'text',
}),
]),
'role': 'assistant',
}),
])
# ---
Reference in New Issue View Git Blame Copy Permalink