Adding support for wildcard models in the model_providers config

[salmanap]

We should be able to support the following config so that developers don't have to restart plano everytime they want to try/use a new model.

listeners:
  # Model listener for direct LLM access
  - type: model
    name: llms
    address: 0.0.0.0
    port: 12000

model_providers:
  # OpenAI - support all models via wildcard
  - model: openai/*
    access_key: $OPENAI_API_KEY

  # Anthropic - support all Claude models
  - model: anthropic/*
    access_key: $ANTHROPIC_API_KEY

  - model: xai/*
    access_key: $GROK_API_KEY


  # Custom internal LLM provider
  # Note: Requires base_url and provider_interface for unknown providers
  - model: ollama/*
    base_url: https://llm.internal.company.com