Lesson 29 — From answer to action: the agent loop
Unit V · Agency

L29_agentkit.py                              student     The agent kit: sandbox help desk, tools with permissions, the loop, three deciders, step log.
L29_run.py                                   student     Runs the T-101 task: scripted hasty (default) or careful, human mode, paste mode.
L29_runs/L29_run_scripted_hasty.txt          student     Captured step log of the hasty run (the traced example).
L29_runs/L29_run_scripted_careful.txt        student     Captured step log of the careful run, for comparison.
L29_trace_worksheet.txt                      student     Annotation worksheet for the traced run, with decision-point proposals.

Python 3.11 or newer. Standard library only. Run each program from the folder the lesson tells you to put it in:
    python <file>.py        (Windows)
    python3 <file>.py       (macOS and Linux)
