Add example script to push traces to otel endpoint by darshana-v · Pull Request #243 · highflame-ai/highflame-python

darshana-v · 2026-01-16T16:52:16Z

No description provided.

darshana-v · 2026-01-16T16:52:25Z

https://github.com/highflame-ai/highflame-core/issues/1864

gemini-code-assist · 2026-01-16T16:52:32Z

Summary of Changes

Hello @darshana-v, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request delivers a new, self-contained example designed to guide developers in instrumenting their AI applications with OpenTelemetry. It provides a practical demonstration of how to capture and export detailed tracing information, particularly for LLM interactions, to an OTLP-compatible monitoring system, thereby enhancing observability for AI-driven services.

Highlights

New Example for OpenTelemetry Traces: Introduced a new example demonstrating how to push OpenTelemetry traces from an AI application to an OTLP endpoint, specifically targeting Highflame Workbench for analysis.
LLM Operation Tracing: The example showcases how to create spans for Large Language Model (LLM) operations (using OpenAI as an example) and enrich them with custom attributes like model, prompts, responses, and token usage metrics.
Comprehensive Documentation: A detailed README.md file has been added, covering prerequisites, installation, environment variable configuration, usage instructions, how to view traces, customization options, and troubleshooting steps.
Python Script for Trace Generation: A Python script (generate_traces.py) is included that initializes an OTLP tracer, performs an OpenAI API call, and meticulously records relevant LLM interaction data as span attributes.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces a valuable example for sending OpenTelemetry traces to Highflame Workbench. The review provides suggestions to enhance the example's clarity, robustness, and adherence to standard practices. Key recommendations include using standard OpenTelemetry environment variables, improving error handling, refactoring hardcoded values, and clarifying the documentation. These changes will make the example more user-friendly and maintainable.

gemini-code-assist · 2026-01-16T16:54:57Z

examples/otel-traces/generate_traces.py

+    resource = Resource.create(
+        {
+            "service.name": os.getenv("OTEL_SERVICE_NAME", "trace-generator"),
+            "service.namespace": "javelin-cerberus",


The service.namespace is hardcoded to "javelin-cerberus". This appears to be an internal or environment-specific value. For a general-purpose example, it's better to remove this line to make the script more broadly applicable without modification.

gemini-code-assist · 2026-01-16T16:54:57Z

examples/otel-traces/generate_traces.py

+
+
+tracer = init_tracer()
+client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])


Accessing os.environ["OPENAI_API_KEY"] directly will raise a KeyError if the environment variable is not set, causing the script to crash. It's more user-friendly to use os.getenv() and check for the variable's existence, raising a descriptive error if it's missing.

api_key = os.getenv("OPENAI_API_KEY") if not api_key: raise ValueError("The OPENAI_API_KEY environment variable is not set.") client = OpenAI(api_key=api_key)

gemini-code-assist · 2026-01-16T16:54:57Z

examples/otel-traces/README.md

+- Example:
+  ```bash
+  export OTEL_EXPORTER_OTLP_HEADERS="your-otel-header"
+  ```


The example value your-otel-header for OTEL_EXPORTER_OTLP_HEADERS is a bit vague. Providing a more concrete example of the expected key=value format would be more helpful for users.

Suggested change

- Example:

```bash

export OTEL_EXPORTER_OTLP_HEADERS="your-otel-header"

```

- Example:

```bash

export OTEL_EXPORTER_OTLP_HEADERS="Authorization=<your-token>"

```

gemini-code-assist · 2026-01-16T16:54:57Z

examples/otel-traces/README.md

+**OTLP_ENDPOINT**
+
+- OTEL endpoint URL
+- Example:
+  ```bash
+  export OTLP_ENDPOINT="https://cerberus-http.api-dev.highflame.dev/v1/traces"
+  ```


The example uses a custom environment variable OTLP_ENDPOINT. To align with OpenTelemetry standards, it's better to use the standard OTEL_EXPORTER_OTLP_TRACES_ENDPOINT variable. This makes the example more familiar to users experienced with OpenTelemetry and allows for simplifying the Python script.

Suggested change

**OTLP_ENDPOINT**

- OTEL endpoint URL

- Example:

```bash

export OTLP_ENDPOINT="https://cerberus-http.api-dev.highflame.dev/v1/traces"

```

**OTEL_EXPORTER_OTLP_TRACES_ENDPOINT**

- OTEL endpoint URL

- Example:

```bash

export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT="https://cerberus-http.api-dev.highflame.dev/v1/traces"

gemini-code-assist · 2026-01-16T16:54:57Z

examples/otel-traces/README.md

+1. Set your environment variables:
+
+```bash
+export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic%20<your-credentials>"


The example for OTEL_EXPORTER_OTLP_HEADERS includes %20, which is likely incorrect. The OpenTelemetry exporter does not URL-decode this value, so %20 will be sent as part of the header. A space should be used instead, with the value quoted in the shell.

Suggested change

export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic%20<your-credentials>"

export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic <your-credentials>"

gemini-code-assist · 2026-01-16T16:54:58Z

examples/otel-traces/generate_traces.py

+def generate_trace() -> None:
+    with tracer.start_as_current_span("openai.chat.completions.create") as span:
+        completion = client.chat.completions.create(
+            model="gpt-4o",


The model name "gpt-4o" is hardcoded here and also when setting span attributes. To improve maintainability and avoid duplication, consider defining it as a constant at the module level and reusing it.

gemini-code-assist · 2026-01-16T16:54:58Z

examples/otel-traces/generate_traces.py

+        )
+
+        span.set_attribute("llm.model", "gpt-4o")
+        span.set_attribute("prompt.user_question", "1 + 1 = ")


The span attribute prompt.user_question is redundant with the input attribute set on line 66. To simplify the example and reduce confusion, it's better to remove this line and use only the input attribute for the prompt.

gemini-code-assist · 2026-01-16T16:54:58Z

examples/otel-traces/generate_traces.py

+        span.set_attribute("llm.model", "gpt-4o")
+        span.set_attribute("prompt.user_question", "1 + 1 = ")
+        span.set_attribute("response.id", completion.id)
+        record_completion_attributes(span, getattr(completion, "usage", {}) or {})


The expression getattr(completion, "usage", {}) or {} is a bit complex. You can make this more readable by passing completion.usage directly to record_completion_attributes. The function already handles cases where usage is None.

record_completion_attributes(span, completion.usage)

gemini-code-assist · 2026-01-16T16:54:58Z

examples/otel-traces/generate_traces.py

+        record_completion_attributes(span, getattr(completion, "usage", {}) or {})
+
+        answer = completion.choices[0].message.content
+        span.set_attribute("response.preview", answer)


The span attribute response.preview is redundant with the output attribute set on line 67. To simplify the example, consider removing this line and using only the output attribute for the response content.

gemini-code-assist · 2026-01-16T16:54:58Z

examples/otel-traces/generate_traces.py

+
+
+if __name__ == "__main__":
+    generate_trace()


The file is missing a final newline character. It's a standard convention to end files with a newline.

generate_trace()

KunalJavelin · 2026-01-17T08:39:07Z

Review Points:

Please address Gemini comments.
This example shows how to push trace to the OTEL endpoint as a standalone file, which is ok but incomplete.
Please use this agent, https://github.com/highflame-ai/highflame-python/tree/main/examples/customer_support_agent, and add steps in there to push traces from a user application to our OTEL endpoint.

Please consider this: if a user has an AI application, what changes will they need to make to push the traces from their application to our endpoint?

darshana-v added 2 commits January 16, 2026 17:47

add generate traces file

a2156b4

add: readme to use the example

abce6a6

darshana-v requested review from KunalJavelin and abhisheksoamj January 16, 2026 16:52

gemini-code-assist bot reviewed Jan 16, 2026

View reviewed changes



		tracer = init_tracer()
		client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

	export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic%20<your-credentials>"
	export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Basic <your-credentials>"



		if __name__ == "__main__":
		generate_trace() No newline at end of file

Conversation

darshana-v commented Jan 16, 2026

Uh oh!

darshana-v commented Jan 16, 2026

Uh oh!

gemini-code-assist bot commented Jan 16, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 16, 2026

Choose a reason for hiding this comment

Uh oh!

KunalJavelin commented Jan 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

KunalJavelin commented Jan 17, 2026 •

edited

Loading