Different parameters for each custom prompt

[wan-robert]

Summary of Feature

Custom prompts currently allow users to specify a model key. It would be very nice if users could also specify model paramters for each prompt.

Why This Is Helpful

Different prompts require different output length and temperatures. A user may want a summary prompt to have very low temperature and short output length. In contrast, a creative writing prompt will benefit from higher temperature and longer output length.

Currently, we have to manually adjust the output length and temperature in the settings each time we want to change the LLM behavior. Allowing users to save parameters with the custom prompt would make this process more efficient for users.

Example

The parameters can be saved with the existing properties, at the head of the markdown file.

---
copilot-command-context-menu-enabled: true
copilot-command-slash-enabled: true
copilot-command-context-menu-order: 160
copilot-command-model-key: gpt-4.1|openai
copilot-command-last-used: 
copilot-command-temperature: 0.8
copilot-command-token-limit: 2000
---

Custom prompt text.

[logancyang]

cc @ichts @zeroliu