Implement Ollama provider for LLM abstraction

[bhouston]

Implement Ollama Provider

We need to add support for the Ollama API to the MyCoder LLM abstraction. Ollama is an open-source framework that allows running large language models locally.

Requirements:

• Create an Ollama provider implementation in packages/agent/src/core/llm/providers/ollama.ts
• Support configurable endpoint URL (default to http://localhost:11434)
• Support both chat and completion API endpoints
• Handle tool/function calling
• Register the provider in the LLM provider registry

Implementation Details:

• The provider should implement the LLMProvider interface
• Support environment variable OLLAMA_BASE_URL for configuration
• Follow similar patterns to existing providers like Anthropic

[bhouston]

This is done here: #212