Skip to content

feat(openai): Gate response output collection on data collection options #7035

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
ericapisani wants to merge 2 commits into
base: master
Choose a base branch
from
+4,118 −3,158
Draft

feat(openai): Gate response output collection on data collection options #7035

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
566ea5c
feat(openai): Gate response output collection on data collection options
ericapisani Aug 4, 2026
770189e
lint
ericapisani Aug 4, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
88 changes: 77 additions & 11 deletions sentry_sdk/integrations/openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -681,12 +681,24 @@ def _set_common_output_data(
integration: "OpenAIIntegration",
finish_span: bool = True,
) -> None:
client = sentry_sdk.get_client()
if hasattr(response, "model"):
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_MODEL, response.model)

# Chat Completions API
if hasattr(response, "choices") and response.choices is not None:
if should_send_default_pii() and integration.include_prompts:
if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
response_text = [
choice.message.model_dump()
for choice in response.choices
if choice.message is not None
]
if len(response_text) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, response_text
)
elif should_send_default_pii() and integration.include_prompts:
response_text = [
choice.message.model_dump()
for choice in response.choices
Expand All @@ -709,12 +721,40 @@ def _set_common_output_data(

# Responses API
elif hasattr(response, "output"):
if should_send_default_pii() and integration.include_prompts:
output_messages: "dict[str, list[Any]]" = {
"response": [],
"tool": [],
}

output_messages: "dict[str, list[Any]]" = {
"response": [],
"tool": [],
}

if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
for output in response.output:
if output.type == "function_call":
output_messages["tool"].append(output.dict())
elif output.type == "message":
for output_message in output.content:
try:
output_messages["response"].append(output_message.text)
except AttributeError:
# Unknown output message type, just return the json
output_messages["response"].append(
output_message.dict()
)

if len(output_messages["tool"]) > 0:
set_data_normalized(
span,
SPANDATA.GEN_AI_RESPONSE_TOOL_CALLS,
output_messages["tool"],
unpack=False,
)

if len(output_messages["response"]) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, output_messages["response"]
)

elif should_send_default_pii() and integration.include_prompts:
for output in response.output:
if output.type == "function_call":
output_messages["tool"].append(output.dict())
Expand Down Expand Up @@ -961,6 +1001,7 @@ def _wrap_synchronous_completions_chunk_iterator(
ttft = None
data_buf: "list[list[str]]" = [] # one for each choice
streaming_message_total_token_usage = None
client = sentry_sdk.get_client()

for x in old_iterator:
if isinstance(span, StreamedSpan):
Expand Down Expand Up @@ -993,7 +1034,12 @@ def _wrap_synchronous_completions_chunk_iterator(
all_responses = None
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
if should_send_default_pii() and integration.include_prompts:
if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses
)
elif should_send_default_pii() and integration.include_prompts:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

_calculate_completions_token_usage(
Expand Down Expand Up @@ -1026,6 +1072,7 @@ async def _wrap_asynchronous_completions_chunk_iterator(
ttft = None
data_buf: "list[list[str]]" = [] # one for each choice
streaming_message_total_token_usage = None
client = sentry_sdk.get_client()

async for x in old_iterator:
if isinstance(span, StreamedSpan):
Expand Down Expand Up @@ -1058,7 +1105,12 @@ async def _wrap_asynchronous_completions_chunk_iterator(
all_responses = None
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
if should_send_default_pii() and integration.include_prompts:
if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses
)
elif should_send_default_pii() and integration.include_prompts:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

_calculate_completions_token_usage(
Expand Down Expand Up @@ -1090,6 +1142,7 @@ def _wrap_synchronous_responses_event_iterator(
"""
ttft = None
data_buf: "list[list[str]]" = [] # one for each choice
client = sentry_sdk.get_client()

count_tokens_manually = True
for x in old_iterator:
Expand Down Expand Up @@ -1125,7 +1178,12 @@ def _wrap_synchronous_responses_event_iterator(
)
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
if should_send_default_pii() and integration.include_prompts:
if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses
)
elif should_send_default_pii() and integration.include_prompts:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

if count_tokens_manually:
Expand Down Expand Up @@ -1157,6 +1215,7 @@ async def _wrap_asynchronous_responses_event_iterator(
"""
ttft: "Optional[float]" = None
data_buf: "list[list[str]]" = [] # one for each choice
client = sentry_sdk.get_client()

count_tokens_manually = True
async for x in old_iterator:
Expand Down Expand Up @@ -1192,8 +1251,15 @@ async def _wrap_asynchronous_responses_event_iterator(
)
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
if should_send_default_pii() and integration.include_prompts:

if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses
)
elif should_send_default_pii() and integration.include_prompts:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

if count_tokens_manually:
_calculate_responses_token_usage(
input=input,
Expand Down
Loading
Loading