test: run every script in smoke, not a 10-of-15 allowlist

[Jammy2211]

Summary

Replaces the smoke_tests.txt allowlist with opt-out coverage: smoke now runs every script under scripts/, minus the exclusions in config/build/no_run.yaml.

Under the allowlist a script was tested only if someone remembered to add it, so every new tutorial was uncovered from birth:

repo	covered	total
HowToGalaxy	4	26
HowToLens	6	40
HowToFit	10	15

That is how a public teaching notebook stayed broken in three places while CI was green (#56 / #57) — no job had ever executed it. The only backstop is PyAutoHeart's workspace-smoke, which runs weekly, and the run that caught it was a manual workflow_dispatch, not the schedule.

Design

run_smoke.py becomes a thin shim over PyAutoHands' autohands/run_python.py — the same entry point PyAutoHeart's workspace-validation uses for its run_scripts job. Discovery, exclusion and env resolution are no longer reimplemented per-repo, so the PR gate and the validation runner cannot drift apart, and both read one exclusion list (config/build/no_run.yaml, which the notebook runner already honoured).

The delegated runner also orders simulator* first and then start_here.py, which is what tutorials depending on simulated datasets need.

--report-dir is passed deliberately, not cosmetically. run_python.py only propagates failures (sys.exit(1)) when a report was built; without it the suite runs to completion and always exits 0 — a vacuously green gate, strictly worse than the allowlist. It also switches execute_script from abort-on-first-failure to record-and-continue, matching the old runner's behaviour. The negative control below is what proves this.

Depends on

PyAutoLabs/PyAutoHands#225 — execute_script() silently rewrote any failure to PASS for scripts with inversion in their path. Without it, chapter_4_pixelizations/tutorial_3_inversions.py would be unconditionally green in this newly-expanded suite, defeating the purpose.

Scripts Changed

• .github/scripts/run_smoke.py — delegates to run_python.py
• smoke_tests.txt — deleted
• AGENTS.md, .gitignore

config/build/no_run.yaml already exists and parses to an empty list, so there is nothing to exclude here and it is unchanged.

Test Plan

• Executed-script count 10 → 15 (report JSON: 15 passed, 0 failed, 0 skipped), EXIT=0
• start_here.py passes in ~0.0s — verified prose-only (0 non-docstring statements), legitimate not vacuous
• Smoke tests pass for all affected workspaces

Generated by the PyAutoLabs agent workflow.