In run_eval.py, when all tasks already have predictions in evals/predictions/,
the function returns early with "No tasks to run" but never calls print_summary().
The user gets no feedback on existing results.
Expected: print existing results summary even when no new tasks run
Actual: exits silently with "No tasks to run"
File: run_eval.py
Function: main
In run_eval.py, when all tasks already have predictions in evals/predictions/,
the function returns early with "No tasks to run" but never calls print_summary().
The user gets no feedback on existing results.
Expected: print existing results summary even when no new tasks run
Actual: exits silently with "No tasks to run"
File: run_eval.py
Function: main