wehub-resource-sync
f3d80b4628
Auto Release / auto-release (push) Failing after 1s
CI / lint (push) Failing after 0s
CI / screenshot-quality (push) Failing after 3s
CI / pr-policy (push) Has been skipped
CI / test (3.11) (push) Failing after 0s
CI / test (3.12) (push) Failing after 0s
CI / test (3.13) (push) Failing after 3s
CI / coverage (push) Failing after 1s
Legibility / legibility (push) Failing after 3s
558 lines
22 KiB
Python
558 lines
22 KiB
Python
"""Focused tests for OpenAI Responses support."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
from pathlib import Path
|
|
|
|
from claude_tap.sse import SSEReassembler
|
|
from claude_tap.viewer import _extract_metadata, _extract_request_messages
|
|
|
|
|
|
def test_sse_reassembler_reconstructs_openai_responses_completed_event() -> None:
|
|
reassembler = SSEReassembler()
|
|
reassembler.feed_bytes(
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"resp_1","status":"in_progress","model":"gpt-5.4"}}\n\n'
|
|
)
|
|
reassembler.feed_bytes(
|
|
b'event: response.completed\ndata: {"type":"response.completed","response":{"id":"resp_1","status":"completed","model":"gpt-5.4","output":[{"type":"message","role":"assistant","content":[{"type":"output_text","text":"Hello!"}]},{"type":"reasoning","summary":[{"type":"summary_text","text":""}]}],"usage":{"input_tokens":12,"output_tokens":3,"reasoning_tokens":0}}}\n\n'
|
|
)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "completed"
|
|
assert body["output"][0]["content"][0]["text"] == "Hello!"
|
|
assert body["usage"] == {"input_tokens": 12, "output_tokens": 3, "reasoning_tokens": 0}
|
|
|
|
|
|
def test_sse_reassembler_recovers_output_from_item_events_when_completed_is_empty() -> None:
|
|
# The Codex/ChatGPT backend over HTTP/SSE (model_provider = "openai_http")
|
|
# streams output via response.output_item.* events and sends a terminal
|
|
# response.completed with output: []. The reassembler must keep the items.
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"resp_1","status":"in_progress","model":"gpt-5.5","output":[]}}\n\n',
|
|
b'event: response.output_item.added\ndata: {"type":"response.output_item.added","output_index":0,"item":{"type":"message","role":"assistant","content":[]}}\n\n',
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":"Hello"}\n\n',
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":" world"}\n\n',
|
|
b'event: response.output_item.done\ndata: {"type":"response.output_item.done","output_index":0,"item":{"type":"message","role":"assistant","content":[{"type":"output_text","text":"Hello world"}]}}\n\n',
|
|
b'event: response.completed\ndata: {"type":"response.completed","response":{"id":"resp_1","status":"completed","model":"gpt-5.5","output":[],"usage":{"input_tokens":13159,"output_tokens":18,"total_tokens":13177}}}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "completed"
|
|
assert body["output"][0]["content"][0]["text"] == "Hello world"
|
|
assert body["usage"] == {"input_tokens": 13159, "output_tokens": 18, "total_tokens": 13177}
|
|
|
|
|
|
def test_sse_reassembler_keeps_streamed_text_when_completed_done_is_missing() -> None:
|
|
# A truncated capture: output_item.done / response.completed never arrive,
|
|
# so the accumulated output_text.delta content must survive on its own.
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"resp_2","status":"in_progress","output":[]}}\n\n',
|
|
b'event: response.output_item.added\ndata: {"type":"response.output_item.added","output_index":0,"item":{"type":"message","role":"assistant","content":[]}}\n\n',
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":"partial"}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["output"][0]["content"][0]["text"] == "partial"
|
|
|
|
|
|
def test_sse_reassembler_merges_response_incomplete_terminal_event() -> None:
|
|
# A response truncated by max_output_tokens ends with response.incomplete,
|
|
# not response.completed. The final status/usage/details must be captured
|
|
# while the streamed output items are preserved.
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","status":"in_progress","output":[]}}\n\n',
|
|
b'event: response.output_item.done\ndata: {"type":"response.output_item.done","output_index":0,"item":{"type":"message","role":"assistant","content":[{"type":"output_text","text":"cut off"}]}}\n\n',
|
|
b'event: response.incomplete\ndata: {"type":"response.incomplete","response":{"id":"r","status":"incomplete","output":[],"incomplete_details":{"reason":"max_output_tokens"},"usage":{"input_tokens":5,"output_tokens":7}}}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "incomplete"
|
|
assert body["incomplete_details"] == {"reason": "max_output_tokens"}
|
|
assert body["usage"] == {"input_tokens": 5, "output_tokens": 7}
|
|
assert body["output"][0]["content"][0]["text"] == "cut off"
|
|
|
|
|
|
def test_sse_reassembler_merges_response_failed_terminal_event() -> None:
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","status":"in_progress","output":[]}}\n\n',
|
|
b'event: response.failed\ndata: {"type":"response.failed","response":{"id":"r","status":"failed","output":[],"error":{"code":"server_error","message":"boom"},"usage":{"input_tokens":3,"output_tokens":0}}}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "failed"
|
|
assert body["error"] == {"code": "server_error", "message": "boom"}
|
|
|
|
|
|
def test_sse_reassembler_records_stream_level_response_error_event() -> None:
|
|
# response.error has no "response" object and can arrive mid-stream; the
|
|
# error must be attached without discarding accumulated output.
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","status":"in_progress","output":[]}}\n\n',
|
|
b'event: response.output_item.done\ndata: {"type":"response.output_item.done","output_index":0,"item":{"type":"message","role":"assistant","content":[{"type":"output_text","text":"hi"}]}}\n\n',
|
|
b'event: response.error\ndata: {"type":"error","code":"rate_limit_exceeded","message":"slow down","param":null}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "failed"
|
|
assert body["error"]["code"] == "rate_limit_exceeded"
|
|
assert body["error"]["message"] == "slow down"
|
|
assert body["output"][0]["content"][0]["text"] == "hi"
|
|
|
|
|
|
def test_sse_reassembler_accepts_output_item_before_response_created() -> None:
|
|
reassembler = SSEReassembler()
|
|
reassembler.feed_bytes(
|
|
b'event: response.output_item.done\ndata: {"type":"response.output_item.done","output_index":"bad","item":{"type":"message","role":"assistant","content":[{"type":"output_text","text":"early"}]}}\n\n'
|
|
)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["output"][0]["content"][0]["text"] == "early"
|
|
|
|
|
|
def test_sse_reassembler_repairs_non_list_responses_output_and_content() -> None:
|
|
reassembler = SSEReassembler()
|
|
for frame in (
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","status":"in_progress","output":{"unexpected":true}}}\n\n',
|
|
b'event: response.output_item.added\ndata: {"type":"response.output_item.added","output_index":0,"item":{"type":"message","role":"assistant","content":{"unexpected":true}}}\n\n',
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","delta":"repaired"}\n\n',
|
|
):
|
|
reassembler.feed_bytes(frame)
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["output"][0]["content"] == [{"type": "output_text", "text": "repaired"}]
|
|
|
|
|
|
def test_sse_reassembler_ignores_malformed_responses_item_and_text_events() -> None:
|
|
item_reassembler = SSEReassembler()
|
|
item_reassembler.feed_bytes(
|
|
b'event: response.output_item.done\ndata: {"type":"response.output_item.done","output_index":0,"item":null}\n\n'
|
|
)
|
|
assert item_reassembler.reconstruct() is None
|
|
|
|
empty_delta_reassembler = SSEReassembler()
|
|
empty_delta_reassembler.feed_bytes(
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":""}\n\n'
|
|
)
|
|
assert empty_delta_reassembler.reconstruct() is None
|
|
|
|
out_of_range_reassembler = SSEReassembler()
|
|
out_of_range_reassembler.feed_bytes(
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","output":[]}}\n\n'
|
|
)
|
|
out_of_range_reassembler.feed_bytes(
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":"ignored"}\n\n'
|
|
)
|
|
assert out_of_range_reassembler.reconstruct() == {"id": "r", "output": []}
|
|
|
|
non_dict_item_reassembler = SSEReassembler()
|
|
non_dict_item_reassembler.feed_bytes(
|
|
b'event: response.created\ndata: {"type":"response.created","response":{"id":"r","output":["raw"]}}\n\n'
|
|
)
|
|
non_dict_item_reassembler.feed_bytes(
|
|
b'event: response.output_text.delta\ndata: {"type":"response.output_text.delta","output_index":0,"delta":"ignored"}\n\n'
|
|
)
|
|
assert non_dict_item_reassembler.reconstruct() == {"id": "r", "output": ["raw"]}
|
|
|
|
|
|
def test_sse_reassembler_handles_responses_terminal_without_response_object() -> None:
|
|
reassembler = SSEReassembler()
|
|
reassembler.feed_bytes(b'event: response.completed\ndata: {"type":"response.completed","id":"r"}\n\n')
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body == {"type": "response.completed", "id": "r"}
|
|
|
|
|
|
def test_sse_reassembler_records_response_error_before_any_snapshot() -> None:
|
|
reassembler = SSEReassembler()
|
|
reassembler.feed_bytes(b'event: response.error\ndata: {"type":"error","code":"server_error","message":"boom"}\n\n')
|
|
|
|
body = reassembler.reconstruct()
|
|
|
|
assert body is not None
|
|
assert body["status"] == "failed"
|
|
assert body["error"] == {"code": "server_error", "message": "boom"}
|
|
|
|
|
|
def test_extract_metadata_supports_responses_input_roles_and_ws_usage() -> None:
|
|
fixture_path = Path(__file__).parent / "fixtures" / "openai_responses_trace.jsonl"
|
|
record_json = fixture_path.read_text(encoding="utf-8").splitlines()[0]
|
|
|
|
meta = _extract_metadata(record_json)
|
|
|
|
assert meta is not None
|
|
assert meta["message_count"] == 1
|
|
assert meta["input_tokens"] == 500
|
|
assert meta["output_tokens"] == 10
|
|
assert meta["model"] == "gpt-5.4"
|
|
assert meta["has_system"] is True
|
|
assert "exec_command" in meta["tool_names"]
|
|
assert "web_search" in meta["tool_names"]
|
|
|
|
|
|
def test_extract_metadata_maps_responses_cached_tokens_to_cache_read() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/responses",
|
|
"body": {"model": "gpt-5.4", "input": [{"role": "user", "content": "hi"}]},
|
|
},
|
|
"response": {
|
|
"status": 200,
|
|
"body": {
|
|
"usage": {
|
|
"input_tokens": 11767,
|
|
"input_tokens_details": {"cached_tokens": 11648},
|
|
"output_tokens": 6,
|
|
"total_tokens": 11773,
|
|
}
|
|
},
|
|
},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["input_tokens"] == 11767
|
|
assert meta["output_tokens"] == 6
|
|
assert meta["cache_read_input_tokens"] == 11648
|
|
|
|
|
|
def test_extract_metadata_falls_back_to_tool_type_and_nested_function_name() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/responses",
|
|
"body": {
|
|
"model": "gpt-5.4",
|
|
"tools": [
|
|
{"type": "tool_search", "description": "# Tool discovery"},
|
|
{
|
|
"type": "function",
|
|
"function": {
|
|
"name": "nested_function",
|
|
"description": "Chat Completions style function tool.",
|
|
},
|
|
},
|
|
],
|
|
},
|
|
},
|
|
"response": {"status": 200, "body": {"usage": {"input_tokens": 1, "output_tokens": 1}}},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert "tool_search" in meta["tool_names"]
|
|
assert "nested_function" in meta["tool_names"]
|
|
|
|
|
|
def test_extract_metadata_counts_responses_tool_search_call_from_body_output() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/responses",
|
|
"body": {"model": "gpt-5.5", "input": [{"role": "user", "content": "Find tools."}]},
|
|
},
|
|
"response": {
|
|
"status": 200,
|
|
"body": {
|
|
"output": [
|
|
{
|
|
"type": "tool_search_call",
|
|
"status": "completed",
|
|
"arguments": {"query": "browser automation", "limit": 5},
|
|
"call_id": "call_search",
|
|
"execution": "client",
|
|
}
|
|
],
|
|
"usage": {"input_tokens": 4, "output_tokens": 2},
|
|
},
|
|
},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["response_tool_names"] == ["tool_search"]
|
|
|
|
|
|
def test_extract_metadata_counts_generic_responses_tool_call_items() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/responses",
|
|
"body": {"model": "gpt-5.5", "input": [{"role": "user", "content": "Search."}]},
|
|
},
|
|
"response": {
|
|
"status": 200,
|
|
"body": {
|
|
"output": [
|
|
{"type": "web_search_call", "status": "completed", "action": {"type": "search", "query": "docs"}},
|
|
{"type": "file_search_call", "status": "completed", "queries": ["parser"]},
|
|
{"type": "custom_tool_call", "status": "completed", "name": "deploy_preview"},
|
|
],
|
|
"usage": {"input_tokens": 4, "output_tokens": 2},
|
|
},
|
|
},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["response_tool_names"] == ["web_search", "file_search", "deploy_preview"]
|
|
|
|
|
|
def test_extract_metadata_counts_ws_tool_search_call_output_item_when_completed_output_is_empty() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "WEBSOCKET",
|
|
"path": "/backend-api/codex/responses",
|
|
"body": {"model": "gpt-5.5", "input": [{"role": "user", "content": "Find tools."}]},
|
|
},
|
|
"response": {
|
|
"status": 101,
|
|
"body": {"output": [], "usage": {"input_tokens": 4, "output_tokens": 2}},
|
|
"ws_events": [
|
|
{"type": "response.created", "response": {"id": "resp_search", "status": "in_progress"}},
|
|
{
|
|
"type": "response.output_item.done",
|
|
"output_index": 0,
|
|
"item": {
|
|
"type": "tool_search_call",
|
|
"status": "completed",
|
|
"arguments": {"query": "browser automation", "limit": 5},
|
|
"call_id": "call_search",
|
|
"execution": "client",
|
|
},
|
|
},
|
|
{
|
|
"type": "response.completed",
|
|
"response": {
|
|
"id": "resp_search",
|
|
"status": "completed",
|
|
"output": [],
|
|
"usage": {"input_tokens": 4, "output_tokens": 2},
|
|
},
|
|
},
|
|
],
|
|
},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["response_tool_names"] == ["tool_search"]
|
|
|
|
|
|
def test_extract_metadata_supports_interleaved_responses_roles_without_type() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request_id": "req_1",
|
|
"timestamp": "2026-03-17T00:00:00Z",
|
|
"duration_ms": 10,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/responses",
|
|
"body": {
|
|
"model": "gpt-5.4",
|
|
"input": [
|
|
{"role": "developer", "content": [{"type": "input_text", "text": "Follow the repo rules."}]},
|
|
{"role": "user", "content": [{"type": "input_text", "text": "Fix the bug."}]},
|
|
{"role": "assistant", "content": [{"type": "output_text", "text": "I will inspect the parser."}]},
|
|
],
|
|
},
|
|
},
|
|
"response": {"status": 200, "body": {"usage": {"input_tokens": 1, "output_tokens": 1}}},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["message_count"] == 3
|
|
assert meta["input_tokens"] == 1
|
|
assert meta["output_tokens"] == 1
|
|
|
|
|
|
def test_extract_metadata_counts_responses_function_call_input_items() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/responses",
|
|
"body": {
|
|
"model": "gpt-5.4",
|
|
"instructions": "Use tools when needed.",
|
|
"input": [
|
|
{"role": "user", "content": [{"type": "input_text", "text": "Read pyproject."}]},
|
|
{
|
|
"type": "function_call",
|
|
"name": "read_file",
|
|
"arguments": '{"path":"pyproject.toml"}',
|
|
},
|
|
{
|
|
"type": "function_call_output",
|
|
"call_id": "call_1",
|
|
"output": '[project]\nname = "claude-tap"',
|
|
},
|
|
],
|
|
"tools": [{"type": "function", "name": "read_file"}],
|
|
},
|
|
},
|
|
"response": {"status": 200, "body": {"usage": {"input_tokens": 4, "output_tokens": 2}}},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["message_count"] == 3
|
|
assert meta["has_system"] is True
|
|
assert meta["input_tokens"] == 4
|
|
assert meta["output_tokens"] == 2
|
|
assert "read_file" in meta["tool_names"]
|
|
|
|
|
|
def test_extract_request_messages_normalizes_responses_function_call_input_items() -> None:
|
|
messages = _extract_request_messages(
|
|
{
|
|
"input": [
|
|
{"role": "user", "content": [{"type": "input_text", "text": "Read pyproject."}]},
|
|
{
|
|
"type": "function_call",
|
|
"name": "read_file",
|
|
"arguments": '{"path":"pyproject.toml"}',
|
|
},
|
|
{
|
|
"type": "function_call_output",
|
|
"call_id": "call_1",
|
|
"output": '[project]\nname = "claude-tap"',
|
|
},
|
|
{
|
|
"type": "function_call",
|
|
"name": "shell",
|
|
"arguments": "not json",
|
|
},
|
|
{
|
|
"type": "function_call",
|
|
"name": "missing_args",
|
|
},
|
|
{
|
|
"type": "web_search_call",
|
|
"action": {"type": "search", "query": "Responses items"},
|
|
},
|
|
{
|
|
"type": "computer_call_output",
|
|
"call_id": "call_screen",
|
|
"output": {"type": "computer_screenshot", "image_url": "https://example.test/screen.png"},
|
|
},
|
|
]
|
|
}
|
|
)
|
|
|
|
assert messages[0]["role"] == "user"
|
|
assert messages[1] == {
|
|
"role": "assistant",
|
|
"content": [{"type": "tool_use", "name": "read_file", "input": {"path": "pyproject.toml"}}],
|
|
}
|
|
assert messages[2] == {"role": "tool", "content": '[project]\nname = "claude-tap"'}
|
|
assert messages[3]["content"][0]["input"] == "not json"
|
|
assert messages[4]["content"][0]["input"] == {}
|
|
assert messages[5] == {
|
|
"role": "assistant",
|
|
"content": [
|
|
{
|
|
"type": "tool_use",
|
|
"name": "web_search",
|
|
"input": {"action": {"type": "search", "query": "Responses items"}},
|
|
}
|
|
],
|
|
}
|
|
assert messages[6]["role"] == "tool"
|
|
assert "computer_screenshot" in messages[6]["content"]
|
|
|
|
|
|
def test_extract_request_messages_normalizes_responses_tool_search_output_input_items() -> None:
|
|
messages = _extract_request_messages(
|
|
{
|
|
"input": [
|
|
{
|
|
"type": "tool_search_output",
|
|
"call_id": "call_search",
|
|
"status": "completed",
|
|
"execution": "client",
|
|
"tools": [
|
|
{
|
|
"type": "namespace",
|
|
"name": "mcp__codex_apps__figma",
|
|
"tools": [{"type": "function", "name": "_use_figma"}],
|
|
}
|
|
],
|
|
}
|
|
]
|
|
}
|
|
)
|
|
|
|
assert len(messages) == 1
|
|
assert messages[0]["role"] == "tool"
|
|
assert "tool_search_output" in messages[0]["content"]
|
|
assert "mcp__codex_apps__figma" in messages[0]["content"]
|
|
assert "mcp__codex_apps__figma._use_figma" in messages[0]["content"]
|
|
|
|
|
|
def test_extract_metadata_ignores_list_response_body() -> None:
|
|
record = {
|
|
"turn": 1,
|
|
"request_id": "req_1",
|
|
"timestamp": "2026-03-17T00:00:00Z",
|
|
"duration_ms": 10,
|
|
"request": {
|
|
"method": "POST",
|
|
"path": "/v1/messages",
|
|
"body": {
|
|
"model": "claude-opus-4-6",
|
|
"messages": [{"role": "user", "content": "hello"}],
|
|
},
|
|
},
|
|
"response": {"status": 200, "body": [{"type": "text", "text": "hello"}]},
|
|
}
|
|
|
|
meta = _extract_metadata(json.dumps(record))
|
|
|
|
assert meta is not None
|
|
assert meta["message_count"] == 1
|
|
assert meta["input_tokens"] == 0
|
|
assert meta["output_tokens"] == 0
|
|
assert meta["error_message"] == ""
|