xbotter/promptplayground

A simple Semantic Kernel semantic function debugging tool.

Prompt Playground

A simple Semantic Kernel prompt debugging tool.

How to use 🐣

Download and Install

1. Download the latest installation package from Release.
2. Install and run prompt-playground.exe.

First-time use

You need to enter the configuration settings for the first time.

Click on the configuration in the menu to enter the configuration interface.

The current model selection supports:

1. Azure OpenAI ChatCompletion (gpt-35/gpt-4)
2. OpenAI ChatCompletion
3. Baidu ERNIE-Bot
4. Aliyun DashScope

Generation quantity refers to the maximum number of results generated, the default is 3.

Fill in the corresponding parameters and close.

Direct use

Enter the prompt directly in the prompt box on the main interface, and click Generate (Ctrl+G/Ctrl+Enter).

Import skprompt

Click the Open File button on the menu bar, select the skprompt.txt file, and the prompt will be automatically imported, and the corresponding config.json file will be imported at the same time.

For skprompts without a config.json file, a default config.json file will be automatically created.

Click the Tab above the input box to switch to the config.json editing interface.

Save skprompt

After modifying the skprompt.txt or config.json file, click the Save (Ctrl+S) button to save to the corresponding file.

Import Semantic Plugin

1. Click on the menu bar, select [Open Folder], and choose the folder containing Semantic Functions
2. Expand the sidebar to see all Semantic Functions in the current folder
3. Select the corresponding Function to edit and run

Screenshots

Build 🛠

1. Install .NET 8 SDK
2. Download the source code
3. Run dotnet build

Roadmap 🚧

See Roadmap

Dependencies 📦

• AvaloniaUI
• Semi.Avalonia
• AvaloniaEdit
• Icons.Avalonia
• Semantic-Kernel
• semantic-kernel-ERNIE-Bot

License 📃

This project is licensed under the MIT License - see the LICENSE file for details

promptplayground's Issues

support handlebars template

识别config.json

• 读取Config.json
• 保存config.json

support function calling

skprompt.txt 文件导入与保存

• skprompt.txt 导入
• skprompt.txt 保存

Bump Semantic Kernel to 1.0.1

批量导入参数并执行

skprompt 参数识别和填充

• skprompt 参数识别
• 参数填充
• #4
• #7

sk Skills项目管理

更多模型支持

• Azure OpenAI
• ERNIE Bot Turbo
• OpenAI
• HuggingFace
• LLama

support yaml file

📍 Roadmap

• #3
• #2
• #8
• Semantic Function Chains
• 历史记录查询
• #5
• #6
• 自动化结果测试
• 多轮结果评估对比
• 配置文件导入导出
• Planner 调试
• i18n
• 技能收藏

生成记录标注

• 添加时间消耗记录
• 添加token消耗记录