feat: add comprehensive mock tests for sequential tool calling

- Created test_sequential_tool_calling.py with 11 test methods
- Added mock infrastructure for realistic LLM response simulation
- Covered various sequential tool calling scenarios
- Support for different LLM providers (OpenAI, Gemini, Ollama, Anthropic)
- Tests for error handling, streaming, and complex arguments

Fixes #847

Co-authored-by: Mervin Praison <[email protected]>

Author: claude[bot]