- Add github.com/ollama/ollama/api dependency (v0.13.5)
- Create OllamaService wrapper with typed handlers for all endpoints
- Implement streaming support for chat, generate, and pull operations
- Add dedicated handlers: ListModels, ShowModel, Chat, Generate, Embed,
PullModel, DeleteModel, CopyModel, Version, Heartbeat
- Keep fallback proxy for any unhandled endpoints
- Update routes to use new typed API endpoints
- Upgrade Go version to 1.24.1
Benefits:
- Type-safe request/response handling
- Built-in NDJSON streaming with 512KB buffer
- Proper tool call types (ToolCall, ToolCallFunction)
- Thinking mode support (ThinkValue)
- Future compatibility with Ollama updates
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>