Mark Wallace

Best practices to control the size of the ChatHistory to avoid exceeding a models maximum context length

5

Here's an example of the type of error a developer can run into ``` Unable to generate bot response. Details: Error: 500: Internal Server ErrorMicrosoft.SemanticKernel.KernelException: StreamResponseToClientAsync failed. ---> Microsoft.SemanticKernel.HttpOperationException: This...

### Motivation and Context This is a POC for the new text search service abstraction. Sharing very early for team collaboration. ### Description ### Contribution Checklist - [ ] The...

.Net: Function Calling Abstractions

1

```[tasklist] ### Tasks - [x] Explore current function-calling model implementation & usage. - [x] POC to understand the scope, feasibility, etc. - [x] Share POC details to have every one...

```[tasklist] ### Tasks - [x] Connector and unit test projects initial check-in - [x] Chat completion support - [x] Embedding support - [x] Streaming chat completion support - [x] Function...

Purpose of this task is to provide some guidance and samples for developers on how control which kernel functions get advertised to the LLM Some options 1. Use the tool...

Guidance on how to debug automatic function calling flows

1

During automatic function calling it can be hard to diagnose issues e.g. if the LLM is selecting the wrong function to call it can be difficult to identify this. The...

Check that case handling of of input variables names is correct and consistent

1

**Describe the bug** A clear and concise description of what the bug is. https://github.com/microsoft/semantic-kernel/pull/6097/files#r1594267057

.Net: Check that handling of input variables with default = "" is correct and consistent

1

**Describe the bug** https://github.com/microsoft/semantic-kernel/pull/6097#discussion_r1594265529

.Net: Support for new ChatHistory serialization formats

1

Provide a way to support new ChatHistory serialization formats i.e. convert a text prompt directly to ChatHistory. This needs to provide protection against prompt injection attacks.

- [ ] Change implementation to use the Azure OpenAI rather than a custom client - [ ] Fix the outstanding issues listed below Chat completion "with data" issues: -...