feat: Add Instructor integration#465
Merged
Merged
Conversation
Abhijeet Prasad (AbhiPrasad)
changed the title
Users can enable Instructor tracing either globally with auto-instrumentation:
```python
import braintrust
braintrust.auto_instrument()
```
or explicitly for Instructor only:
```python
from braintrust.integrations import InstructorIntegration
InstructorIntegration.setup()
```
Manual wrapping is also supported for direct client/class wrapping:
```python
from braintrust import wrap_instructor
client = wrap_instructor(client)
```
The integration instruments sync and async Instructor structured-output APIs, including `create`, `create_with_completion`, `create_partial`, and `create_iterable`.
It emits `task` spans for Instructor structured-output work. Each span captures the response model, mode, messages, max retries, retry count, validation errors, and the extracted Pydantic output (or collected iterable/partial outputs):
```python
{
"input": {
"response_model": "Person",
"mode": "TOOLS",
"messages": [...],
},
"output": {"name": "Grace", "age": 45},
"metadata": {
"response_model": "Person",
"mode": "TOOLS",
"max_retries": 3,
"retry_count": 0,
"validation_errors": [],
},
}
```
Token usage and LLM request/response spans remain owned by the underlying provider integrations such as OpenAI or Anthropic to avoid double-counting.
Luca Forstner (lforst)
approved these changes
May 26, 2026
Abhijeet Prasad (AbhiPrasad)
force-pushed
the
feat/instructor-integration
branch
from
fb7c031 to
b3d0117
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
ref https://python.useinstructor.com/
resolves #460
Users can enable Instructor tracing either globally with auto-instrumentation:
or explicitly for Instructor only:
Manual wrapping is also supported for direct client/class wrapping:
The integration instruments sync and async Instructor structured-output APIs, including
create,create_with_completion,create_partial, andcreate_iterable.It emits
taskspans for Instructor structured-output work. Each span captures the response model, mode, messages, max retries, retry count, validation errors, and the extracted Pydantic output (or collected iterable/partial outputs):{ "input": { "response_model": "Person", "mode": "TOOLS", "messages": [...], }, "output": {"name": "Grace", "age": 45}, "metadata": { "response_model": "Person", "mode": "TOOLS", "max_retries": 3, "retry_count": 0, "validation_errors": [], }, }Token usage and LLM request/response spans remain owned by the underlying provider integrations such as OpenAI or Anthropic to avoid double-counting.