coalex 0.4.0

OpenTelemetry SDK for Coalex.ai with OTLP integration

Coalex SDK

A Python package for OpenTelemetry integration with Coalex.ai observability platform.

Installation

pip install coalex

Quick Start with VertexAI

from coalex.otel import register, coalex_context
from openinference.instrumentation.vertexai import VertexAIInstrumentor
from vertexai.generative_models import GenerativeModel

# Register Coalex tracing
tracer_provider = register(
    agent_id="YOUR_AGENT_ID"
)

# Instrument VertexAI with Coalex tracer
VertexAIInstrumentor().instrument(tracer_provider=tracer_provider)

# Create model instance
model = GenerativeModel("gemini-2.0-flash")

# Use coalex_context for proper span management and attribute propagation
with coalex_context(
    request_id="req_vertex_001",
    prompt_version="v1.0.0"
):
    # Generate content - this will be automatically traced with proper context
    response = model.generate_content(
        "Write a haiku about artificial intelligence.",
        generation_config={
            "max_output_tokens": 100,
            "temperature": 0.7,
        }
    )

print("Generated haiku:")
print(response.text)

Features

• Simple Setup: One-line registration with register()
• Coalex Integration: Default endpoint for Coalex observability platform
• Context Management: Use coalex_context() for proper span hierarchy and attribute propagation
• OpenInference Compatible: Works with all OpenInference instrumentors
• Authentication: Automatic authentication using agent_id

Configuration

• agent_id: Your unique agent identifier (required for authentication)
• endpoint: OTLP endpoint (defaults to Coalex: https://traces.coalex.ai/v1/traces)

Example

See examples/vertexai_example.py for a complete working example including:

• Basic content generation
• Streaming responses
• Proper error handling
• Context management with coalex_context()

To run the example, install the additional dependencies:

pip install google-cloud-aiplatform openinference-instrumentation-vertexai