feat: streamline pre-launch setup and implement interactive in-chat slash commands

[Nithwin]

Is your feature request related to a problem? Please describe.

Having two distinct sets of commands (CLI subcommands like windmist set ... vs in-chat interaction) or having to configure API keys and models before starting the AI session creates an annoying two-step workflow. Furthermore, inside the active chat loop (windmist chat), commands like /model and /provider currently return "coming soon." instead of letting users switch settings mid-conversation.

Describe the solution you'd like

1. Keep CLI subcommands lightweight (windmist set ...) for scripting/initial setup, but make them optional rather than mandatory pre-requisites.
2. Activate In-Chat Slash Commands: When inside the interactive WindMist chat session (windmist chat), typing /provider or /model will trigger the arrow-key selection overlay right inside the chat UI!
3. Seamless Mid-Chat Switching: Selecting a new provider or model via /provider or /model will instantly update the active AI agent's configuration on the fly and confirm with a chat bubble ("✨ Switched to Ollama (model: qwen2.5:8b)") without having to exit and restart the application.
4. If an API key is missing when a provider is selected (either at startup or via /provider), prompt the user right inside the terminal to enter it (/api_key).

Describe alternatives you've considered

No response

Additional Context

No response