[Ygormorais]

• I have added tests that cover my changes.
• If adding a new instrumentation or changing an existing one, I've added screenshots from some observability platform showing the change. (Bug fix only: the reproduction output and unit tests are below.)
• PR name follows conventional commits format: feat(instrumentation): ... or fix(instrumentation): ....
• (If applicable) I have updated the documentation accordingly. (Not applicable.)

What

_set_input_attributes in the watsonx instrumentation uses SpanAttributes.GEN_AI_PROMPT when the prompt is a list:

f"{SpanAttributes.GEN_AI_PROMPT}.{index}.user"

opentelemetry.semconv_ai.SpanAttributes has no GEN_AI_PROMPT member (not in the published 0.5.1, not in the in-repo version), so this raises AttributeError. _handle_input is wrapped in @dont_throw, so the error is swallowed. The result:

• list prompts (model.generate(prompt=["a", "b"])) are silently missing from the span
• _emit_input_events is never reached for these calls, because it runs after the failing line

This PR uses GenAIAttributes.GEN_AI_PROMPT (already imported in the module, and the documented replacement for SpanAttributes.LLM_PROMPTS in semconv-ai's MIGRATION.md) for the list branch, and for the string branch too, which used LLM_PROMPTS. The attribute key stays gen_ai.prompt, so the string-prompt output is unchanged.

I checked the latest main and the 150 most recent PR heads: none changes this line.

Reproduction (before the fix)

string error=None  prompt attrs={'gen_ai.prompt.0.user': 'hello'}
list   error=AttributeError: type object 'SpanAttributes' has no attribute 'GEN_AI_PROMPT'  prompt attrs={}

Tests

Added tests/traces/test_input_attributes.py (no cassettes, no network): string prompt, list prompt passed as kwarg, list prompt passed positionally.

• without the fix: 2 failed, 1 passed
• with the fix: full watsonx suite passes (11 passed); ruff check . clean

🤖 Generated with Claude Code

https://claude.ai/code/session_019vQj79oorBrk3jGJ9eumPM

Summary by CodeRabbit

• Improvements
  • Prompt telemetry now uses standardized GenAI attribute names for scalar and list inputs.
  • List prompts continue to retain indexed user attributes, including when provided positionally.

[CLAassistant]

All committers have signed the CLA.

[coderabbitai]

No actionable comments were generated in the recent review.

Walkthrough

Watsonx prompt attributes now use the GenAI prompt namespace for scalar and list inputs. Tests cover trimmed string prompts, list prompts, and positional list prompts.

Changes: Watsonx prompt attributes

Layer / File(s)	Summary
Prompt keys and tests packages/opentelemetry-instrumentation-watsonx/opentelemetry/instrumentation/watsonx/__init__.py, packages/opentelemetry-instrumentation-watsonx/tests/traces/test_input_attributes.py	Scalar and list prompts use indexed gen_ai.prompt user attributes. Tests cover trimmed scalar prompts and keyword and positional lists.

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Suggested reviewers: huang-cn

Merge Risk: Minimal · up to 67447

This fixes list prompts being silently dropped from watsonx spans, and the change is small and covered by new tests. No merge-blocking risk was found.

Pre-merge checks: 4

Check name	Status	Explanation
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.
Title check		The title names the fix (list prompts recorded with GenAIAttributes.GEN_AI_PROMPT in the watsonx instrumentation). The raw summary confirms the change swaps the prompt attribute namespace for scalar a…
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.

Full details: Title check

Explanation

The title names the fix (list prompts recorded with GenAIAttributes.GEN_AI_PROMPT in the watsonx instrumentation). The raw summary confirms the change swaps the prompt attribute namespace for scalar and list inputs, and the objectives describe replacing the nonexistent SpanAttributes.GEN_AI_PROMPT member. The title matches the main change, so it passes.

Finishing Touches

Generate unit tests (beta)

• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Recent review info

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: 25668b20-98d4-41c3-9de3-b44b5a5ab8a2

Commits

Reviewing files that changed from the base of the PR and between e3d651a and 6744714.

Files selected for processing: 2

• packages/opentelemetry-instrumentation-watsonx/opentelemetry/instrumentation/watsonx/__init__.py
• packages/opentelemetry-instrumentation-watsonx/tests/traces/test_input_attributes.py

Included review availability: This review used your included allowance. Your plan provides up to 8 included reviews per hour; 7 remain after this review.