Getting Started
Prerequisites
- Python 3.9 or newer
pip
Source code: Codon SDK on GitHub
Platform Setup
To access the Codon observability platform and export telemetry data, you'll need to create an account and obtain an API key.
Step 1: Access the Login Screen
Navigate to the Codon platform and click 'Sign in with Github':

Step 2: Authorize with GitHub
You'll be redirected to GitHub to authorize the Codon application:

Step 3: Access Your Dashboard
After authorization, you'll be redirected to your organization dashboard:

Your dashboard will display your organization ID, name, email, and API key.
Step 4: Configure Your Organization
Add or update your organization name and save your settings:

Important: Copy your API key and set it as the CODON_API_KEY environment variable to authenticate telemetry exports to the Codon platform.
Installation
Install the Codon SDK from PyPI:
pip install codon-sdk
For framework-specific integrations, install the corresponding instrumentation packages:
# LangGraph integration
pip install codon-instrumentation-langgraph
Environment Configuration
| Variable | Purpose |
|---|---|
ORG_NAMESPACE |
Required by NodeSpec and instrumentation to scope identifiers. |
CODON_API_KEY |
Required to export telemetry data to the Codon observability platform. |
OTEL_SERVICE_NAME |
Default service name for telemetry (overridden by service_name parameter). |
OTEL_EXPORTER_OTLP_ENDPOINT |
Configure custom OTLP collector endpoint. |
Set ORG_NAMESPACE before constructing NodeSpec objects or instrumented decorators will raise a validation error.
You can set these environment variables directly:
export ORG_NAMESPACE=your-org-name
export OTEL_SERVICE_NAME=your-service-name # default service name
export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4317
export CODON_API_KEY=your-api-key-from-dashboard # required for telemetry on Codon platform
Or create a .env file in your project root:
# Required
ORG_NAMESPACE=your-org-name
# Optional - default service name (overridden by function parameter)
OTEL_SERVICE_NAME=your-service-name
OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4317
# Required for captured telemetry visible on Codon platform
CODON_API_KEY=your-api-key-from-dashboard
Initializing Telemetry
Once you have your API key configured, initialize telemetry to start sending observability data to the Codon platform:
from codon_sdk.instrumentation import initialize_telemetry
# Initialize telemetry - uses CODON_API_KEY automatically
initialize_telemetry()
Call initialize_telemetry() once at the start of your application, before creating workloads or executing agents. This function:
- Configures OpenTelemetry to export spans to the configured endpoint (via
OTEL_EXPORTER_OTLP_ENDPOINT) - Automatically uses your
CODON_API_KEYenvironment variable for authentication (can be overridden withapi_keyparameter) - Uses
OTEL_SERVICE_NAMEfor service identification, with function parameter taking precedence - Works for both from-scratch workloads and framework integrations
Example with service name:
initialize_telemetry(service_name="my-ai-agent")
Next Steps
Now that you have the SDK installed and configured, you can:
- Build from scratch: Create custom agents with CodonWorkload
- Use existing frameworks: Integrate with LangGraph or other supported frameworks
- Learn the APIs: Explore detailed documentation in the API Reference
For detailed information about NodeSpec and Logic ID generation, see the API Reference.