Parent
Part of #34
Goal
Verify tool execution paths including code generation, web search, and sandbox restrictions.
What to test
- Sandbox capability restrictions block disallowed operations
- Function CRUD (create, update, delete, retrieve)
- Web search request formation and response handling
- Code generation validation rejects unsafe or invalid output
- Sandbox timeout enforcement
Parent
Part of #34
Goal
Verify tool execution paths including code generation, web search, and sandbox restrictions.
What to test