fix(anthropic): record streaming latency before safety masking - #43
Merged
sayonfortify merged 1 commit intoJun 20, 2026
Merged
Conversation
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fixes Anthropic TTFT/STTG span attributes when FortifyRoot streaming safety masks or buffers the first text token. The wrapper now detects the first streaming token before applying streaming-safety processing, then continues the existing safety/pending-item flow.
Why
Live dev telemetry showed Anthropic streaming usage events landing in RDS, but TTFT/STTG stayed NULL when a config profile was enabled. A local spy confirmed the latency attributes were skipped because safety processing could mutate the token event before
_is_streaming_token_item(...)ran.Validation
python -m py_compileon the Anthropic streaming wrapper and test filepython -m pytest tests/test_safety_unit.py -q -k "streaming_latency or latency_attrs_before_safety"-> 5 passedpv-demo-streaming-latency-anthropicNotes
No provider API behavior changes; only span attribute timing changes. Histograms and safety masking behavior remain unchanged.