feat: Kimi/Moonshot partnership promotion POC

[felarof99]

Summary

• Add Moonshot / Kimi as the first quick provider template in LLM Providers settings, pre-configured with Moonshot API (https://api.moonshot.ai/v1, model kimi-k2-0905-preview)
• Add "Free model Sponsored by Kimi K2.5" promotional badge on the BrowserOS built-in provider card
• Add Kimi at the top of LLM Chat & Hub with orange glow ring and "Try Kimi" badge
• Register moonshot as a recognized provider type across shared schema, all 3 server provider factories, and frontend

Design

Moonshot API is OpenAI-compatible, so the server uses createOpenAICompatible for the factory — no new adapter code. Frontend changes add moonshot to the ProviderType union, provider templates, icons (using Kimi from @lobehub/icons), models data, and form schema. The promotional styling uses existing --accent-orange CSS variables and Tailwind utilities.

13 files changed across packages/shared, apps/server, and apps/agent.

Test plan

• Build the agent extension and verify "Moonshot / Kimi" appears as first template in LLM Providers settings
• Click "USE" on the Kimi template → verify form pre-fills with Moonshot API URL and kimi-k2-0905-preview model
• Verify "Free model Sponsored by Kimi K2.5" badge shows on the BrowserOS provider card
• Navigate to LLM Chat & Hub → verify Kimi is at the top with orange glow and "Try Kimi" badge
• Test provider connection with a valid Moonshot API key
• Server typecheck passes (bun run --filter @browseros/server typecheck)

🤖 Generated with Claude Code

[greptile-apps]

Greptile Summary

This PR adds Moonshot/Kimi as a partnership promotion across the BrowserOS agent. The implementation consistently adds moonshot as a new provider type across all layers (shared schemas, server factories, and frontend). Moonshot uses OpenAI-compatible API, so no new adapter code is needed.

Key changes:

• Moonshot/Kimi positioned as first provider template in LLM settings
• "Free model Sponsored by Kimi K2.5" badge on BrowserOS provider card
• Kimi promoted at top of LLM Hub with orange glow ring and "Try Kimi" badge
• Server implementations added to all 3 provider factory locations (provider-adapter, tool-loop, llm client)

Minor concern:

• LLM Hub uses hardcoded name check for promotional styling, which is fragile if users edit the provider name

Confidence Score: 4/5

• This PR is safe to merge with minimal risk
• Implementation is straightforward and consistent across all files. The moonshot provider reuses existing OpenAI-compatible adapter pattern, minimizing risk. Minor style concern with hardcoded name check in LLM Hub doesn't affect core functionality.
• No files require special attention - all changes follow established patterns

Important Files Changed

Filename	Overview
apps/agent/entrypoints/app/llm-hub/HubProviderRow.tsx	Adds Kimi promotional styling (orange glow ring and "Try Kimi" badge) using hardcoded name check
apps/agent/entrypoints/app/ai-settings/ProviderCard.tsx	Adds sponsorship badge to BrowserOS provider card with correct type check
apps/agent/lib/llm-providers/providerTemplates.ts	Added Moonshot/Kimi as first provider template with API URL and default model
packages/shared/src/schemas/llm.ts	Added moonshot to LLM provider schema constants and enum
apps/server/src/agent/provider-adapter/index.ts	Added createMoonshotFactory using OpenAI-compatible adapter

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[User opens LLM Settings] --> B{Selects Provider}
    B -->|New Provider| C[NewProviderDialog]
    C --> D[moonshot first in providerTypeEnum]
    C --> E[Select Moonshot/Kimi template]
    E --> F[Pre-fills API URL & model]
    
    B -->|Existing Provider| G[ProviderCard]
    G -->|BrowserOS provider| H{Check provider.type}
    H -->|type === 'browseros'| I[Show Kimi K2.5 sponsorship badge]
    
    J[User opens LLM Hub] --> K[Load DEFAULT_PROVIDERS]
    K --> L[Kimi at top of list]
    L --> M[HubProviderRow]
    M -->|Check provider.name| N{name === 'Kimi'?}
    N -->|Yes| O[Apply orange glow + Try Kimi badge]
    N -->|No| P[Standard styling]
    
    Q[Server receives request] --> R{Provider type?}
    R -->|moonshot| S[createMoonshotFactory]
    S --> T[createOpenAICompatible]
    T --> U[Vercel AI SDK]
    
    style D fill:#ff9800
    style I fill:#ff9800
    style O fill:#ff9800
    style S fill:#ff9800

Loading

_{Last reviewed commit: 7579aff}

[greptile-apps]

_{13 files reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}

[greptile-apps]

Badge shows for ALL built-in providers, but PR description says it should only show "on the BrowserOS built-in provider card". Need to check provider.type === 'browseros' instead of just isBuiltIn.

Suggested change

	{isBuiltIn && (
	<span className="mb-1 inline-block rounded-full bg-gradient-to-r from-orange-500 to-amber-500 px-3 py-0.5 font-semibold text-white text-xs">
	Free model Sponsored by Kimi K2.5
	</span>
	)}
	{isBuiltIn && provider.type === 'browseros' && (
	<span className="mb-1 inline-block rounded-full bg-gradient-to-r from-orange-500 to-amber-500 px-3 py-0.5 font-semibold text-white text-xs">
	Free model Sponsored by Kimi K2.5
	</span>
	)}

Prompt To Fix With AI

This is a comment left during a code review.
Path: apps/agent/entrypoints/app/ai-settings/ProviderCard.tsx
Line: 80-84

Comment:
Badge shows for ALL built-in providers, but PR description says it should only show "on the BrowserOS built-in provider card". Need to check `provider.type === 'browseros'` instead of just `isBuiltIn`.

```suggestion
        {isBuiltIn && provider.type === 'browseros' && (
          <span className="mb-1 inline-block rounded-full bg-gradient-to-r from-orange-500 to-amber-500 px-3 py-0.5 font-semibold text-white text-xs">
            Free model Sponsored by Kimi K2.5
          </span>
        )}
```

How can I resolve this? If you propose a fix, please make it concise.