Feature Request: API Key Model Whitelisting/Blacklisting

Summary: Add the ability to restrict API keys to only work with specific models (whitelist) or exclude certain models (blacklist).

Use Case 1: Parental Controls / Family Safety / SFW

As a parent setting up Agent0 for my child, I want to create an API key that:

• Blocks access to: Lustify and other uncensored/adult image/video generation models

• Allows access to: Safe text models and SFW image generation only

This would let me give my son API access for coding, homework help, and creative projects without worrying about accidental (or sought after) exposure to adult content. Also applies to the workplace.

Use Case 2: Cost Control / Budget Protection

I want to create "safe" API keys that:

• Exclude expensive models: Claude Opus, GPT-4.5, and other high-cost models

• Allow affordable models: Llama, Qwen, Mistral, etc.

This prevents accidental API calls that could rack up unexpected costs, especially when testing or sharing keys with team members.