Skip to content

ref(openai): Drop send_default_pii support #7733

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

Merged
alexander-alderman-webb merged 1 commit into from
Sep 29, 2026
Merged

ref(openai): Drop send_default_pii support #7733

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
156 changes: 26 additions & 130 deletions sentry_sdk/integrations/openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,12 +41,10 @@
)
from sentry_sdk.consts import SPANDATA
from sentry_sdk.integrations import DidNotEnable, Integration, _check_minimum_version
from sentry_sdk.scope import should_send_default_pii
from sentry_sdk.traces import Span
from sentry_sdk.utils import (
capture_internal_exceptions,
event_from_exception,
has_data_collection_enabled,
parse_version,
reraise,
)
Expand Down Expand Up @@ -369,33 +367,16 @@ def _set_responses_api_input_data(
reasoning["effort"],
)

client_options = sentry_sdk.get_client().options
if has_data_collection_enabled(client_options):
if client_options["data_collection"]["gen_ai"]["inputs"]:
tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_responses(tools)),
)
else:
# Pre-data collection this was always set, so this needs to be left here for now until
# we deprecate `send_default_pii`. Once we do, this 'else' branch should be removed,
# and the above branch placed below the "if not should_send_default_pii()"
# line below
tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_responses(tools)),
)

if has_data_collection_enabled(client_options):
if not client_options["data_collection"]["gen_ai"]["inputs"]:
return
elif not should_send_default_pii():
if not sentry_sdk.get_client().options["data_collection"]["gen_ai"]["inputs"]:
return

tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_responses(tools)),
)

explicit_instructions: "Union[Optional[str], Omit]" = kwargs.get("instructions")
has_explicit_instructions = explicit_instructions is not None and _is_given(
explicit_instructions
Expand Down Expand Up @@ -497,37 +478,20 @@ def _set_completions_api_input_data(
if reasoning_level is not None and _is_given(reasoning_level):
span.set_attribute(SPANDATA.GEN_AI_REQUEST_REASONING_LEVEL, reasoning_level)

client = sentry_sdk.get_client()
if has_data_collection_enabled(client.options):
if client.options["data_collection"]["gen_ai"]["inputs"]:
tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_completions(tools)),
)
else:
# Pre-data collection this was always set, so this needs to be left here for now until
# we deprecate `send_default_pii`. Once we do, this 'else' branch should be removed,
# and the above branch placed below the "if not should_send_default_pii()"
# line below
tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_completions(tools)),
)
if not sentry_sdk.get_client().options["data_collection"]["gen_ai"]["inputs"]:
return

tools = kwargs.get("tools")
if tools is not None and _is_given(tools):
span.set_attribute(
SPANDATA.GEN_AI_TOOL_DEFINITIONS,
json.dumps(_transform_tool_definitions_completions(tools)),
)

messages: "Optional[Union[str, Iterable[ChatCompletionMessageParam]]]" = kwargs.get(
"messages"
)

if has_data_collection_enabled(client.options):
if not client.options["data_collection"]["gen_ai"]["inputs"]:
return
elif not should_send_default_pii():
return

if messages is None:
return

Expand Down Expand Up @@ -562,7 +526,6 @@ def _set_completions_api_input_data(
]
if len(non_system_messages) > 0:
normalized_messages = normalize_message_roles(non_system_messages) # type: ignore
client = sentry_sdk.get_client()
set_data_normalized(
span,
SPANDATA.GEN_AI_REQUEST_MESSAGES,
Expand All @@ -583,17 +546,13 @@ def _set_embeddings_input_data(
if model is not None:
span.set_attribute(SPANDATA.GEN_AI_REQUEST_MODEL, model)

if not sentry_sdk.get_client().options["data_collection"]["gen_ai"]["inputs"]:
return

messages: "Optional[Union[str, SequenceNotStr[str], Iterable[int], Iterable[Iterable[int]]]]" = kwargs.get(
"input"
)

client = sentry_sdk.get_client()
if has_data_collection_enabled(client.options):
if not client.options["data_collection"]["gen_ai"]["inputs"]:
return
elif not should_send_default_pii():
return

if messages is None:
return

Expand Down Expand Up @@ -640,18 +599,7 @@ def _set_common_output_data(

# Chat Completions API
if hasattr(response, "choices") and response.choices is not None:
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():
if client.options["data_collection"]["gen_ai"]["outputs"]:
response_text = [
choice.message.model_dump()
for choice in response.choices
Expand Down Expand Up @@ -681,39 +629,7 @@ def _set_common_output_data(
"tool": [],
}

if has_data_collection_enabled(client.options) and isinstance(
response.output, list
):
record_outputs = client.options["data_collection"]["gen_ai"]["outputs"]

if record_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) # type: ignore[union-attr]
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 record_outputs and len(output_messages["response"]) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_RESPONSE_TEXT, output_messages["response"]
)

elif should_send_default_pii() and isinstance(response.output, list):
if client.options["data_collection"]["gen_ai"]["outputs"]:
for output in response.output:
if output.type == "function_call":
output_messages["tool"].append(output.dict())
Expand Down Expand Up @@ -983,12 +899,7 @@ def _wrap_synchronous_completions_chunk_iterator(
all_responses = None
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
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():
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

_calculate_completions_token_usage(
Expand Down Expand Up @@ -1051,12 +962,7 @@ 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 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():
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

_calculate_completions_token_usage(
Expand Down Expand Up @@ -1121,12 +1027,7 @@ def _wrap_synchronous_responses_event_iterator(
)
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]
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():
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

if count_tokens_manually:
Expand Down Expand Up @@ -1192,12 +1093,7 @@ async def _wrap_asynchronous_responses_event_iterator(
if len(data_buf) > 0:
all_responses = ["".join(chunk) for chunk in data_buf]

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():
if client.options["data_collection"]["gen_ai"]["outputs"]:
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_TEXT, all_responses)

if count_tokens_manually:
Expand Down
Loading
Loading