A collection of useful functions to be deployed as custom skills for Azure Cognitive Search
A collection of useful functions to be deployed as custom skills for Azure Cognitive Search
[!IMPORTANT]
As of November 15, 2023, Azure Cognitive Search has been renamed to Azure AI Search.
A collection of sample implementations for custom skill functionality in Azure AI Search.
Previously known as: Azure Search Power Skills
This repository may still be referenced by its former name in older documentation, links, and conversations.
Terminology clarification
In this repository, “skills” refers to Azure AI Search skillset enrichment skills (including Custom Web API skills used during indexing), not agent/copilot skills.
Azure AI Search includes many built-in enrichment skills.
This repository provides examples for scenarios where you need custom logic that built-in skills don’t provide directly, such as:
This repo contains reusable samples and patterns for:
Language composition of this repository:
Use built-in skills when:
Use custom skills when:
Start here for official guidance:
Custom Web API skill
https://learn.microsoft.com/azure/search/cognitive-search-custom-skill-web-api
Create and define skillsets
https://learn.microsoft.com/azure/search/cognitive-search-defining-skillset
Built-in skills reference
https://learn.microsoft.com/azure/search/cognitive-search-predefined-skills
Indexers in Azure AI Search
https://learn.microsoft.com/azure/search/search-indexer-overview
Enrichment concepts (skillsets/pipeline context)
https://learn.microsoft.com/azure/search/cognitive-search-concept-intro
This repository is specifically for Azure AI Search skillset custom enrichment samples and related implementation patterns.
This project provides the following custom skills:
| Skill | Description | Type | Language | Environment | Deployment |
|---|---|---|---|---|---|
| Chat completion Custom Skill | Demonstrates how to use chat-completion inference capabilities of language models (including Azure OpenAI models) deployed in Azure AI Foundry in a custom web API skillset in Azure AI Search. | Text | Azure Functions | ||
| GeoPointFromName | Retrieves coordinates from place names and addresses. | Geography | Azure Functions | ||
| Distinct | De-duplicates a list of terms. | Text | Azure Functions | ||
| TextAnalyticsForHealth | A wrapper for the Text Analytics for Health API. | Text | Azure Functions | ||
| TextQualityWatchdog | Uses a pretrained language model to detect low-quality text extracted during document cracking. | Text | Azure Function and Azure Open AI endpoint | ||
| DecryptBlobFile | Downloads, decrypts, and returns a file previously encrypted and stored in Azure Blob Storage. | Utility | Azure Functions | ||
| GetFileExtension | Returns the filename and extension as separate values to allow filtering on document type. | Utility | Azure Functions | ||
| HelloWorld | A minimal skill that can be used as a starting point or template for your own skills. | Template | Azure Functions | ||
| PythonFastAPI | A production web server and API scaffold for a Python power skill. | Template | Azure Functions | ||
| VideoIndexer | Extracts insights from video content for indexing workflows. | Vision | Azure Functions |
In order to use the functions in this project, you'll need an active Azure subscription. Most of the functions can be used on their own for quick evaluation and experimentation, but they are meant for integration in Azure AI Search enrichment pipelines.
Each function may also add its own specific requirements, such as API keys for services they leverage.
Visual Studio is recommended, but not required. You need a recent version of the C# compiler.
If using Visual Studio with the Azure workload installed, no installation is required, and the functions can just be run locally using F5.
Deployment of a function to Azure can be done through Visual Studio, the Deploy to Azure button, or continuous deployment from source control.
Some functions may require setting environment variables or configuration entries. Please refer to the readme file in the function's directory.
You can also create your own skills using our Hello World template skill as a starting point or if you are using python our FastAPI template skill.
Here are a few suggestions of simple contributions to get you started:
Text/, Utils/, Template/) as references for clearer setup guidance and consistent config naming.Video/VideoIndexer or Text/TextAnalyticsForHealth), consider adding deployment templates and setup instructions to simplify onboarding and reproducibility.No open issues yet, or sync has not completed.