Feature request: Support Command Code (commandcode.ai) as a configurable model provider

[paulwalker9315-debug]

Feature Request: Add Command Code (commandcode.ai) as a model provider

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

I use Command Code as my coding agent. It offers flat-rate subscriptions with 2-10x cheaper model access than direct API prices (DeepSeek, GPT-5.6, Qwen, MiniMax, Kimi, etc.) and exposes a Provider API that is compatible with both OpenAI and Anthropic endpoints. I'd like to configure Command Code models inside Proma so I can use its credits/subscription from Proma's model settings, instead of paying full API prices elsewhere.

Describe the solution you'd like

Add Command Code as a configurable model provider in Proma:

• Provider base URL: https://commandcode.ai/provider
• OpenAI-compatible endpoint: /chat/completions (for OpenAI and open-source models)
• Anthropic-compatible endpoint: /messages (for Claude / Anthropic models)
• Auth: Bearer token from a Command Code account (Provider API plan)
• Allow selecting model IDs listed at https://commandcode.ai/docs/reference/cli/models (e.g., DeepSeek / GPT-5.6 / Qwen / Kimi / GLM / MiniMax family)

Describe alternatives you've considered

• Use a generic "custom OpenAI-compatible provider" config as a workaround (base URL + API key), which should already cover /chat/completions models.
• Use the Anthropic-compatible endpoint to call Claude models through Command Code.

Additional context

• Product: https://commandcode.ai
• Pricing: https://commandcode.ai/pricing
• Provider API docs: https://commandcode.ai/docs/provider
• CLI model list: https://commandcode.ai/docs/reference/cli/models