Use python3 as the canonical command name course-wide (#104)

Most current systems (default Debian/Ubuntu, recent macOS) install Python
only as `python3`, with no bare `python` on PATH, so learners who copied
`python cli.py ...` into their host shell hit "command not found".

Convert host-shell `python <cmd>` -> `python3 <cmd>` across module/lab
READMEs, lab `.py` docstrings & usage strings, blog posts, lab prompt and
instruction files, the M04 verify.sh message, and the M10/M24 lab patches.
Module 01's convention note (and its blog/02 mirror) is rewritten so
`python3` is canonical and `python` is the documented fallback.

Stop-lines respected: Docker image tags (`python:3.12-slim`), `.venv/.../python`
and `...\.venv\Scripts\python.exe` paths, the M20 `"command": "python"`
teaching example and surrounding venv prose, container-internal invocations
(M16/M18 Dockerfiles, M16 README `docker run` examples), and CI-workflow
`run:` steps fed by `actions/setup-python` / `image: python:3.12` are left
as `python` on purpose.

pip was left out of scope: most occurrences are prose or CI/container-internal,
and `pip3` does not fix the PEP 668 externally-managed-environment refusal that
the course already addresses with venvs. The M01 note is worded to stay
consistent with bare `pip` (use whichever pip pairs with your Python).

Build (tools/build_wiki.py) and tools/check.sh both pass.

Closes #104

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GAEzanEoGJT5o1VizQar47

blog/16-unit5-ai-in-the-loop.md

@@ -40,9 +40,9 @@ The lab makes this concrete and local: no hosted bot account required. You run a
 
 ```bash
 cd modules/24-assistive-agents/lab
-python reviewer.py prompt          # builds: your committed rubric + the diff
+python3 reviewer.py prompt          # builds: your committed rubric + the diff
 # (paste into your AI, save its JSON to my-review.json)
-python reviewer.py apply my-review.json
+python3 reviewer.py apply my-review.json
 ```
 
 The diff it's reviewing has a real trap planted in it: a new `clear` command that prints "cleared all tasks" but never actually calls `save()`, so `tasks.json` is untouched. Did your AI catch it? Either way, *you* make the merge call, and you learn exactly how much this reviewer is worth before the stakes go up.
@@ -70,7 +70,7 @@ The lab runs the whole thing locally against the `tasks-app`, and the best part
 
 ```bash
 git checkout -b agent/delete-command
-python agent_runner.py issue-to-pr issue-delete-command.md --simulate bad
+python3 agent_runner.py issue-to-pr issue-delete-command.md --simulate bad
 # → ruff + pytest run, a test fails, the script refuses to call the work ready.
 #   Exit code non-zero. No PR. Nothing reached main.
 ```
@@ -124,8 +124,8 @@ The lab is the punchline of the whole series. You run the same eval set against
 
 ```bash
 cd modules/27-evals/lab
-python run_eval.py candidates/current_model   # 100%, exit 0, your baseline
-python run_eval.py candidates/swapped_model    # 60%, exit 1, blocked
+python3 run_eval.py candidates/current_model   # 100%, exit 0, your baseline
+python3 run_eval.py candidates/swapped_model    # 60%, exit 1, blocked
 ```
 
 The "swapped model" is a stand-in for the day a cheaper model ships, or your provider deprecates the one you're on, or someone edits the agent's prompt. The easy cases still pass (this output would sail through a casual skim), but the eval caught a regression a skim would have missed, *and the non-zero exit code means a pipeline would have blocked the merge.* That's a **regression eval**, and it's the moment this course's thesis stops being a slogan and becomes a procedure you run from the keyboard.