Getting to know the core plugins
Large language models are great at generating text and images, but they are currently unreliable for several other tasks, for example:
- Retrieving data from external data sources, such as the filesystem and the web
- Interacting with the system’s clock
- Performing math, even simple arithmetic functions
The Semantic Kernel installation includes several plugins for tasks that are very frequently used. The core plugins are native plugins that provide functionality that LLMs struggle with.
To use the plugins in C#, you need to import the Microsoft.SemanticKernelCoreSkills library, and to use the plugins in Python, you need to import the plugins from the semantic_kernel.core_skills library.
Skills or plugins?
When the Microsoft Semantic Kernel was first released, plugins were called skills. This changed when OpenAI allowed developers to build extensions to ChatGPT and called these extensions plugins. Since the purpose is...
