g3/g3-plan/completed_todo_2025-12-09_22-43-24.md

Planner Mode UI and Error Handling Refinements

1. Error Propagation from LLM Calls

• Add error handling to call_refinement_llm_with_tools() in crates/g3-planner/src/llm.rs
  • Import classify_error and ErrorType from g3_core::error_handling
  • Wrap agent execution with error classification
  • Display user-friendly error messages based on error type

2. Single-Line Tool Output Display

• Modify print_tool_header() in PlannerUiWriter to accept tool arguments
  • Change signature to accept tool_args: Option<&serde_json::Value>
  • Format output as single line with first 50 chars of args
  • Ensure no trailing newlines
  • Update UiWriter trait and all implementations
  • Update call site in g3-core to pass tool args

3. Display LLM Text Responses

• Fix print_agent_response() to prevent overwriting
  • Use println instead of print to avoid overwriting
  • Review notify_sse_received() for carriage return issues
  • Update print_status_line() to use proper formatting

4. Consistent Workspace Logs Directory

• Set G3_WORKSPACE_PATH early in run_planning_mode()
  • Move env var setting before provider initialization
  • Create logs directory and verify it exists
  • Add user notification about logs directory
  • Remove duplicate G3_WORKSPACE_PATH setting in coach_player_loop

Testing

• Test error display with rate limit scenario
• Test tool output formatting
• Test text response visibility
• Verify logs are written to workspace/logs directory

Summary

All implementations complete and verified:

• Error handling with classify_error() properly integrated
• Tool output displays on single line with args preview
• Text responses use println to avoid overwrites
• Workspace path set early, logs directory created consistently
• Code compiles successfully with no errors