Extension Decorator¶

A meta-decorator that enables loading of community-defined decorators from external sources. This facilitates the use of specialized decorator packages, domain-specific extensions, or custom decorator libraries maintained by communities or organizations.

Category: Meta

Parameters¶

Parameter	Type	Description	Default
`source`	`string`	URI or identifier for the extension package (e.g., URL, namespace, or registry identifier)	`Required`
`version`	`string`	Specific version of the extension package to use	``
`decorators`	`array`	Specific decorators to load from the extension (if empty, loads all decorators from the package)	``

Examples¶

Basic loading of an extension package¶

+++Extension(source=https://decorator-registry.ai/scientific-pack.json)
+++ScientificReasoning(discipline=physics)
Explain dark matter.

Loads decorators from the scientific-pack extension and then applies the ScientificReasoning decorator (defined in that pack) with physics discipline to explain dark matter

Loading specific decorators from a versioned extension¶

+++Extension(source=medical-decorators, version=2.1.0, decorators=[ClinicalCase,EvidenceBased])
+++ClinicalCase(format=SOAP)
Describe the treatment approach for Type 2 diabetes.

Loads only the ClinicalCase and EvidenceBased decorators from version 2.1.0 of the medical-decorators package, then applies the ClinicalCase decorator with SOAP format to describe diabetes treatment

Model-Specific Implementations¶

gpt-4o¶

Instruction: This request will load specialized decorators from {source}. If any decorators in the following prompt come from this extension, apply them according to their definitions in the extension package.

Notes: This model handles extension loading well but requires clear indications of which decorators come from the extension

Implementation Guidance¶

Loading scientific package for dark matter explanation¶

Original Prompt:

+++Extension(source=https://decorator-registry.ai/scientific-pack.json)
+++ScientificReasoning(discipline=physics)
Explain dark matter.

Transformed Prompt:

Please load and apply specialized decorators from an external source to enhance your response capabilities. Load the extension package from this source: https://decorator-registry.ai/scientific-pack.json.

+++ScientificReasoning(discipline=physics)
Explain dark matter.

Loading specific medical decorators for diabetes treatment¶

Original Prompt:

+++Extension(source=medical-decorators, version=2.1.0, decorators=[ClinicalCase,EvidenceBased])
+++ClinicalCase(format=SOAP)
Describe the treatment approach for Type 2 diabetes.

Transformed Prompt:

Please load and apply specialized decorators from an external source to enhance your response capabilities. Load the extension package from this source: medical-decorators. Use version 2.1.0 of the extension package. Load only these specific decorators from the package: ClinicalCase,EvidenceBased.

+++ClinicalCase(format=SOAP)
Describe the treatment approach for Type 2 diabetes.

Transformation Details¶

Base Instruction: Please load and apply specialized decorators from an external source to enhance your response capabilities.

Placement: prepend

Composition Behavior: override

Parameter Effects:

source:
Format: Load the extension package from this source: {value}.
version:
Format: Use version {value} of the extension package.
decorators:
Format: Load only these specific decorators from the package: {value}.

Compatibility¶

Requires: None
Conflicts: None
Compatible Models: gpt-4o
Standard Version: 1.0.0 - 2.0.0

Context: Enhances Extension Context can help adapt the loaded extension decorators to domain-specific requirements
Custom: Enhances Extension Custom provides flexibility for working with extension decorators that might require customization
Version: Enhances Extension Version helps ensure compatibility between core decorators and loaded extensions