Implement a code-evaluation harness that runs unit tests against generated solutions and r

Medium

Agents: Tool Use and Post-Training4 / 7

Chapter 194 / 7

Agents: Tool Use and Post-Training

Medium

Implement a code-evaluation harness that runs unit tests against generated solutions and returns pass/fail.

Target interface

evaluate_solution(solution_code, test_code, timeout_s=5.0)

Complete the starter code in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.

solution.pyPython

Local draft

import subprocess, tempfile, os, json

def evaluate_solution(solution_code, test_code, timeout_s=5.0):
    raise NotImplementedError

Local draft

⌘/Ctrl↵to submit

AI review

Ready when you are

Submit your solution for a score and feedback on correctness, edge cases, and clarity. Any correct approach is accepted.

Chapter 194 / 7

Agents: Tool Use and Post-Training

Medium

Implement a code-evaluation harness that runs unit tests against generated solutions and returns pass/fail.

Target interface

evaluate_solution(solution_code, test_code, timeout_s=5.0)

Complete the starter code in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.