Add LiteLLM provider support

[AnishDe12020]

Add native support for LiteLLM Proxy as a model provider. LiteLLM provides a unified OpenAI-compatible API gateway for 100+ LLM providers (OpenAI, Anthropic, Gemini, etc.) with features like load balancing, fallbacks, and spend tracking.

Current Behavior

When configuring LiteLLM as a custom provider, Spacebot sends model names with the provider prefix (e.g., litellm/kimi-k2p5), but LiteLLM expects plain model IDs (kimi-k2p5). This results in 400 errors:
400: {'error': '/chat/completions: Invalid model name passed in model=litellm/kimi-k2p5'}

Proposed Solution

Add LiteLLM to the remap_model_name_for_api() function in src/llm/model.rs to strip the provider prefix:

fn remap_model_name_for_api(provider: &str, model_name: &str) -> String {
    if provider == "zai-coding-plan" {
        // Coding Plan endpoint expects plain model ids (e.g. "glm-5").
        model_name
            .strip_prefix("zai/")
            .unwrap_or(model_name)
            .to_string()
    } else if provider == "litellm" {
        // LiteLLM expects plain model ids (e.g. "kimi-k2p5").
        model_name
            .strip_prefix("litellm/")
            .unwrap_or(model_name)
            .to_string()
    } else {
        model_name.to_string()
    }
}