feat: add support for NVIDIA NIM provider in LLM system

[Aaryan Bansal (NotUnHackable)]

Add NVIDIA NIM AI API Support
Summary
Adds support for NVIDIA NIM (NVIDIA Inference Microservices) as a new LLM provider in BrowserOS.
Changes
Files modified:

• packages/shared/src/schemas/llm.ts - Added nvidia-nim provider constant and schema
• packages/shared/src/constants/urls.ts - Added NVIDIA NIM API endpoint
• apps/agent/lib/llm-providers/types.ts - Added nvidia-nim to ProviderType
• apps/agent/lib/llm-providers/providerTemplates.ts - Added NVIDIA NIM template
• apps/agent/lib/llm-providers/providerIcons.tsx - Added icon mapping
• apps/server/src/lib/clients/llm/provider.ts - Added provider factory
• apps/server/src/agent/provider-factory.ts - Added agent factory
  Usage
• Default Base URL: https://integrate.api.nvidia.com/v1
• Default Model: nvidia/llama-3.1-nemotron-nano-8b-instruct
• API Keys: Get at https://build.nvidia.com/settings/api-keys
• Docs: https://docs.nvidia.com/nim/
  Notes
• Uses OpenAI-compatible API (via @ai-sdk/openai-compatible)
• Free tier: 1,000 inference credits on signup (40 RPM rate limit)
• Supports 100+ models including Nemotron, Kimi, MiniMax, GLM

[greptile-apps]

PR author is not in the allowed authors list.

[github-actions]

All contributors have signed the CLA. Thank you!
_{Posted by the CLA Assistant Lite bot.}

[Aaryan Bansal (NotUnHackable)]

I have read the CLA Document and I hereby sign the CLA

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.