callforAgentTesting

Test Addresses

Mac ARM Test Address
Mac x86 Test Address
Linux Test Address

Test Objectives

1. How many languages can DeepSeek support for translation?
   

2. The robustness of the current system prompt.
   https://github.com/SamYuan1990/i18n-agent-action/blob/main/Business/translateConfig.py#L67-L89

3. Consistency of development frameworks like Flet across platforms (Mac, Linux). Personally, I believe future AI Agents will support various integration methods. Therefore, if frameworks like Flet can enable cross-platform compilation and building from a single codebase to deliver a consistent user experience, it would be a great choice.

Test Steps and Scope

My personal computer is a Mac x86, so I will use it as a reference.

1. Download and install the software.

You may encounter trust issues with the signature. Try a few times if needed.

1. Configure a DeepSeek API key.
   Please refer to https://api-docs.deepseek.com/zh-cn/ or create one via the web platform.
   

Of course, everyone is also welcome to use their existing OpenAI-format large language models to expand the testing scope.

1. Configure the large language model access information and save it.
   

2. Enter the content to be translated and click "Translate" to wait for the result (note: voice output is enabled by default).
   

3. Optional feature: Reserved words.
   Back to Step 1, adding reserved word and reproduce to step 4.

If you have technical skills and would like to try speech-to-text, please contact me. Currently, due to technical limitations, only a development version is available.