Skip to content

fix(pydantic-ai): Stop capturing tool execution spans when validation fails #7448

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 2 commits into from
Sep 10, 2026
+12 −35
Merged

fix(pydantic-ai): Stop capturing tool execution spans when validation fails #7448

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
44f1ddb
fix(pydantic-ai): Stop capturing tool execution spans when validation…
alexander-alderman-webb Sep 9, 2026
1fc035d
.
alexander-alderman-webb Sep 9, 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
18 changes: 12 additions & 6 deletions sentry_sdk/integrations/pydantic_ai/patches/tools.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
import sys
from contextlib import nullcontext
from functools import wraps
from typing import TYPE_CHECKING

Expand Down Expand Up @@ -60,11 +61,15 @@ async def wrapped_execute_tool_call(
# Create execute_tool span
# Nesting is handled by isolation_scope() to ensure proper parent-child relationships
with sentry_sdk.isolation_scope():
with execute_tool_span(
name,
args_dict,
agent,
tool_definition=selected_tool_definition,
with (
execute_tool_span(
name,
args_dict,
agent,
tool_definition=selected_tool_definition,
)
if validated.args_valid
else nullcontext()
) as span:
try:
result = await original_execute_tool_call(
Expand All @@ -73,7 +78,8 @@ async def wrapped_execute_tool_call(
*args,
**kwargs,
)
update_execute_tool_span(span, result)
if span is not None:
update_execute_tool_span(span, result)
return result
except ToolRetryError as exc:
exc_info = sys.exc_info()
Expand Down
29 changes: 0 additions & 29 deletions tests/integrations/pydantic_ai/test_pydantic_ai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1022,24 +1022,6 @@ def add_numbers(a: Annotated[int, Field(gt=0, lt=0)], b: int) -> int:
chat_spans = [
s for s in spans if s["attributes"].get("sentry.op", "") == "gen_ai.chat"
]
tool_spans = [
s
for s in spans
if s["attributes"].get("sentry.op", "") == "gen_ai.execute_tool"
]

# Should have tool spans
assert len(tool_spans) >= 1

# Check tool spans
model_retry_tool_span = tool_spans[0]
assert "execute_tool" in model_retry_tool_span["name"]
assert (
model_retry_tool_span["attributes"]["gen_ai.operation.name"]
== "execute_tool"
)
assert model_retry_tool_span["attributes"]["gen_ai.tool.name"] == "add_numbers"
assert "gen_ai.tool.input" in model_retry_tool_span["attributes"]

# Check chat spans have available_tools
assert "gen_ai.request.available_tools" in chat_spans[0]["attributes"]
Expand Down Expand Up @@ -1073,17 +1055,6 @@ def add_numbers(a: Annotated[int, Field(gt=0, lt=0)], b: int) -> int:

# Find child span types (invoke_agent is the transaction, not a child span)
chat_spans = [s for s in spans if s["op"] == "gen_ai.chat"]
tool_spans = [s for s in spans if s["op"] == "gen_ai.execute_tool"]

# Should have tool spans
assert len(tool_spans) >= 1

# Check tool spans
model_retry_tool_span = tool_spans[0]
assert "execute_tool" in model_retry_tool_span["description"]
assert model_retry_tool_span["data"]["gen_ai.operation.name"] == "execute_tool"
assert model_retry_tool_span["data"]["gen_ai.tool.name"] == "add_numbers"
assert "gen_ai.tool.input" in model_retry_tool_span["data"]

# Check chat spans have available_tools
assert "gen_ai.request.available_tools" in chat_spans[0]["data"]
Expand Down
Loading