feat: bootstrap HowToGalaxy repo with tutorial content and CI

[Jammy2211]

Summary

Extracts the HowToGalaxy tutorial series from autogalaxy_workspace/howtogalaxy/ into its own dedicated repository at PyAutoLabs/HowToGalaxy, mirroring the HowToLens pattern.

This is sub-task 1 of 3 in the HowToGalaxy extraction:

1. [this PR] Bootstrap PyAutoLabs/HowToGalaxy with tutorial content + CI scaffolding
2. Clean howtogalaxy/ out of autogalaxy_workspace (separate PR)
3. Update PyAutoGalaxy docs / README / Colab URLs to point at the new repo (separate PR)

What's included

Tutorial content (ported from autogalaxy_workspace/howtogalaxy/)

• scripts/ — all 5 chapters plus simulators/:
  • chapter_1_introduction/ — grids, galaxies, data, fitting, methods, summary
  • chapter_2_modeling/ — non-linear search, galaxy modeling
  • chapter_3_search_chaining/ — search chaining, prior passing, pipelines
  • chapter_4_pixelizations/ — pixelized reconstructions, inversions, regularization
  • chapter_optional/ — alternative non-linear searches
  • simulators/ — simulator scripts that generate tutorial datasets at runtime
• notebooks/ — matching .ipynb versions (regenerated by PyAutoBuild downstream)
• config/ — full PyAutoGalaxy config tree (from autogalaxy_workspace/config/)

Path prefix scripts/howtogalaxy/ → scripts/ (and same for notebooks/) rewritten via sed across all files.

Scaffolding (derived from HowToLens, rewritten for PyAutoGalaxy)

• .github/workflows/smoke_tests.yml — runs .github/scripts/run_smoke.py against smoke_tests.txt (without the PyAutoLens checkout/install step HowToLens has)
• .github/workflows/url_check.yml — same link-checking workflow as HowToLens
• .github/scripts/run_smoke.py — verbatim from HowToLens
• welcome.py — rewritten for import autogalaxy as ag, HowToGalaxy working-directory checks
• start_here.py — rewritten for HowToGalaxy's chapter layout (chapter_2_modeling, not chapter_2_lens_modeling)
• README.rst, CLAUDE.md — HowToGalaxy-specific (references PyAutoGalaxy, autogalaxy_workspace, and notes the PyAutoBuild target still needs adding)
• CITATIONS.rst — replaced with PyAutoGalaxy's version
• CODE_OF_CONDUCT.md, CONTRIBUTING.md, runtime.txt, .gitignore, LICENSE — from HowToLens
• requirements.txt — autogalaxy\nnumba
• version.txt — 2026.4.13.6 (matches the current autogalaxy_workspace release; tagged in this PR)
• smoke_tests.txt — chapter-1 tutorials 1, 2, 4, 5

Smoke test status (verified locally, PYAUTO_TEST_MODE=2)

Tutorial	Status
tutorial_1_grids_and_galaxies.py	✅ PASS
tutorial_2_data.py	✅ PASS
tutorial_4_methods.py	✅ PASS
tutorial_5_summary.py	✅ PASS

Intentionally excluded from smoke_tests.txt

• tutorial_0_visualization.py — inherits an upstream bug (uses Path(...) without importing) and references dataset/imaging/simple__sersic with no auto-sim fallback
• tutorial_3_fitting.py — auto-sim block calls scripts/imaging/simulator.py (does not exist in HowToGalaxy); a second dataset load for dataset/imaging/simple has no auto-sim fallback

Both are tracked as a content-alignment follow-up to this bootstrap PR.

Known follow-ups

• Register howtogalaxy target in PyAutoBuild/autobuild/config.yaml so notebooks regenerate automatically (same pattern as the HowToLens follow-up).
• Fix tutorial_0 imports and add auto-sim blocks to tutorials 0 and 3 so they can join smoke_tests.txt.

Test plan

• Chapter-1 smoke tests pass locally under PYAUTO_TEST_MODE=2
• CI smoke_tests.yml workflow passes on this PR
• CI url_check.yml workflow passes on this PR

Closes PyAutoLabs/autogalaxy_workspace#35 (sub-task 1 of 3).

🤖 Generated with Claude Code