fix: resolve TypeError when passing non-iterable ToolNode objects to langchain instrumentation tools extraction

[manojmanivannan]

• 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.
• PR name follows conventional commits format: feat(instrumentation): ... or fix(instrumentation): ....
• (If applicable) I have updated the documentation accordingly.

Summary by CodeRabbit

Release Notes

• Bug Fixes

  • Enhanced tool definition extraction in LangChain instrumentation to support alternative tool configurations.

• Tests

  • Added test coverage for ToolNode-based tool handling in agent creation.

[CLAassistant]

All committers have signed the CLA.

[coderabbitai]

📝 Walkthrough

Walkthrough

The changes extend the agent wrapper's tool-definition extraction to conditionally iterate over a tools_by_name attribute when present, falling back to direct iteration otherwise, and introduce a test verifying this behavior with ToolNode objects passed to create_react_agent.

Changes

Cohort / File(s)	Summary
Tool Definition Extraction Enhancement packages/opentelemetry-instrumentation-langchain/opentelemetry/instrumentation/langchain/patch.py	Modified create_agent_wrapper to detect and iterate over tools_by_name.values() when available on the tools object, with fallback to direct tools iteration.
Agent Creation Test Addition packages/opentelemetry-instrumentation-langchain/tests/test_langgraph.py	Added test_create_react_agent_with_toolnode to verify GenAI semantic convention attributes (operation name, agent name, tool definitions) when ToolNode is passed to create_react_agent.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐰 A tool by any name shall still be found,
Whether in a list or nested deep,
Our agent now checks all around,
Tool definitions we shall keep! 🔧

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 55.56% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main fix: handling non-iterable ToolNode objects in langchain tools extraction by detecting and iterating over the tools_by_name attribute.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

packages/opentelemetry-instrumentation-langchain/tests/test_langgraph.py (1)

558-591: Test adequately covers the fix; consider strengthening the assertion.

The test validates that the create_agent span is produced with GEN_AI_TOOL_DEFINITIONS when a ToolNode is passed — which is exactly the regression case. To make the test guard against future regressions where tools_by_name iteration silently yields zero definitions, consider additionally asserting the tool name is present in the serialized definitions:

✅ Optional stronger assertion

     assert create_span.attributes[GenAIAttributes.GEN_AI_AGENT_NAME] == "ToolNodeAgent"
     assert GenAIAttributes.GEN_AI_TOOL_DEFINITIONS in create_span.attributes
+    import json as _json
+    tool_defs = _json.loads(create_span.attributes[GenAIAttributes.GEN_AI_TOOL_DEFINITIONS])
+    assert any(t.get("name") == "get_weather" for t in tool_defs)

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@packages/opentelemetry-instrumentation-langchain/tests/test_langgraph.py`
around lines 558 - 591, The test should additionally assert that the serialized
tool definitions actually include the tool's name to prevent silent regressions;
in test_create_react_agent_with_toolnode after locating create_span (from
create_react_agent with ToolNode) deserialize or inspect
create_span.attributes[GenAIAttributes.GEN_AI_TOOL_DEFINITIONS] and assert it
contains "get_weather" (the tool defined via `@tool`) so the ToolNode ->
tools_by_name serialization is validated.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@packages/opentelemetry-instrumentation-langchain/tests/test_langgraph.py`:
- Around line 558-591: The test should additionally assert that the serialized
tool definitions actually include the tool's name to prevent silent regressions;
in test_create_react_agent_with_toolnode after locating create_span (from
create_react_agent with ToolNode) deserialize or inspect
create_span.attributes[GenAIAttributes.GEN_AI_TOOL_DEFINITIONS] and assert it
contains "get_weather" (the tool defined via `@tool`) so the ToolNode ->
tools_by_name serialization is validated.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 43da2013-62c3-4925-9846-f1bb140d44fe

📥 Commits

Reviewing files that changed from the base of the PR and between 3efa0eb and 4f12775.

📒 Files selected for processing (2)

• packages/opentelemetry-instrumentation-langchain/opentelemetry/instrumentation/langchain/patch.py
• packages/opentelemetry-instrumentation-langchain/tests/test_langgraph.py