Reflexion reports success rates across repeated trials on the same tasks, showing how much reflections improve the agent over attempts.
Implement success_rate(results), the fraction of successful attempts.
Count the truthy results and divide by the total.
No results yet
Run your code to check it against the examples, or submit to run every test.