docs: record 2026-07-26 regime-split round in child prompts

draft/docs/autolens/docs_three_regime_restructure.md

@@ -7,7 +7,7 @@ Repos:
 Difficulty: medium
 Autonomy: supervised
 Priority: high
-Status: draft
+Status: in progress — all planned edits landed 2026-07-26 (branch claude/pyautolens-doc-reorganization-w6a1l5); pending RTD build confirmation on merge
 Parent: draft/docs/autolens/split_lensing_regimes.md
 
 Restructure the PyAutoLens Sphinx/RTD documentation (`PyAutoLens/docs/`) around
@@ -19,7 +19,31 @@ example links, API pointers and modelling philosophy.
 
 - `docs/overview/overview_2_new_user_guide.md`: the multi_galaxy rung and the
   four-step ladder routing with the analysis-split prose landed (PyAutoLens
-  `c086863` + review fixes). The remaining items below are still open.
+  `c086863` + review fixes). `overview_3_features.md`: group section renamed
+  "Multi-Galaxy Lenses, Groups and Clusters" (PyAutoLens#653).
+
+## Landed (2026-07-26, this task branch)
+
+- `docs/overview/overview_1_start_here.md`: the three regime enumerations
+  now spell out the four-rung ladder (galaxy-scale, multi-galaxy,
+  group-scale, cluster-scale).
+- `docs/api/mass.rst`: regime note on the Total section — untruncated
+  Isothermal/PowerLaw for galaxy/multi_galaxy, tidally truncated dPIE
+  members for group/cluster (sigma_lt vs b0 parameterizations named).
+- `docs/api/point.rst`: intro states point sources are the standard source
+  strategy at cluster scale (per-source redshifts, multi-plane) with the
+  workspace `cluster` package pointer.
+- `docs/api/galaxy.rst`: new "Galaxy Catalogues (CSV)" section documenting
+  `galaxy_table_from_csv` / `galaxies_from_csv_tables` /
+  `galaxy_af_models_from_csv_tables` (previously absent from the API docs)
+  with the scaling-tier regime framing.
+- `docs/general/model_cookbook.md`: new "Multi Galaxy, Group and Cluster
+  Models" section — per-deflector loop recipe, shared-prior scaling-relation
+  tie recipe, cluster point-source framing, links to the four regime
+  notebooks.
+
+This completes the planned scope: the API reference stays single-copy with
+per-surface regime notes (as specified), rather than a three-way fork.
 
 ## Changes
 
@@ -33,22 +57,13 @@ example links, API pointers and modelling philosophy.
     point-source/multi-image-position workflow by default.
   Include the taxonomy sentence: all groups and clusters are multi-galaxy
   systems, but not vice versa.
-- `docs/overview/overview_1_start_here.md` + `overview_3_features.md`: add the
-  regime split where scales are enumerated; link the three start_here Colab
-  notebooks.
-- `docs/api/`: ensure the mass/galaxy/point API pages surface the
-  regime-relevant surfaces where users will look for them — dPIE profiles,
-  `al.sr` scaling relations, `galaxy_table_from_csv` /
-  `galaxies_from_csv_tables` / `galaxy_models_from_csv` CSV APIs, and the
-  point-source/`PointSolver` machinery — grouped or cross-referenced by
-  regime (a short "which regime uses this" note per surface is enough; do not
-  fork the API reference into three copies).
-- `docs/general/model_cookbook.md`: add multi_galaxy, group (both
-  with/without-halo compositions) and cluster model recipes.
-- Scientific grounding: each regime section cites 2–3 flagship
-  systems/surveys from the parent plan's literature research (e.g. the
-  multi_galaxy flagship, SL2S/CASSOWARY groups, HFF/A2744 clusters) so the
-  docs point at real, recognisable science.
+- ~~`docs/overview/overview_1_start_here.md` + `overview_3_features.md`~~
+  LANDED (see above).
+- ~~`docs/api/`~~ LANDED (see above; per-surface regime notes, no fork).
+- ~~`docs/general/model_cookbook.md`~~ LANDED (see above).
+- ~~Scientific grounding~~ LANDED 2026-07-26: the New User Guide ladder now
+  cites J1011+0143/B1608+656 (multi_galaxy), CSWA 19/SL2S (group) and
+  HFF/A2744 (cluster).
 
 ## Ordering
 

draft/docs/autolens/multi_galaxy_package.md

@@ -8,7 +8,7 @@ Repos:
 Difficulty: large
 Autonomy: supervised
 Priority: high
-Status: in progress — core landed on branch claude/pyautolens-doc-reorganization-w6a1l5 (2026-07-25)
+Status: in progress — core landed 2026-07-25; features/fit/jax legs landed 2026-07-26 (branch claude/pyautolens-doc-reorganization-w6a1l5); only the real-data swap-in remains
 Parent: draft/docs/autolens/split_lensing_regimes.md
 
 ## Landed (2026-07-25, this task branch)
@@ -26,15 +26,36 @@ Parent: draft/docs/autolens/split_lensing_regimes.md
 - PyAutoLens docs: New User Guide four-rung ladder + multi_galaxy links
   (full RTD restructure remains with docs_three_regime_restructure.md).
 
+## Landed (2026-07-26, this task branch)
+
+- `features/scaling_galaxies/` (simulator + modeling): five faint galaxies
+  4–7" out on an untruncated-`IsothermalSph` relation
+  (`einstein_radius = einstein_radius_ref * (L/L_ref)**0.5`, truth ref
+  0.15"), shared-prior tier costing one free parameter, the "load of
+  galaxies far from the lens / not a standard ingredient" framing, and the
+  group/cluster truncation contrast. Smoke-registered and validated green.
+- `fit.py`: N-deflector fit anatomy — per-galaxy deflection fields compared
+  (mean-|deflection| co-dominance ratio 0.82), truth-composition
+  `FitImaging` (full-res truth LL +27220), pointing at `imaging/fit.py` for
+  the step-by-step API anatomy. Smoke-registered and validated green.
+  Scope decision: a `likelihood_function.py` mirror was deliberately NOT
+  added — the likelihood machinery is regime-independent and fully
+  documented by `imaging/` + the group package; `fit.py` covers the one
+  regime-specific piece (summed deflection fields).
+- autolens_workspace_test `multi_galaxy/jax_likelihood/lp.py`: batched
+  `fitness._vmap` literal + `jit(fit_from)` NumPy-parity round-trip over a
+  two-co-dominant-deflector model; smoke-registered, runner-validated
+  (15s).
+
 ## Remaining
 
-- features/ scripts (extra_galaxies, scaling_galaxies with untruncated
-  isothermals, pixelization) — currently README cross-links only.
 - Swap start_here to the REAL SDSS J1011+0143 HST data (F555W/F814W via
   MAST) once frames are prepared; the simulated look-alike is the interim.
-- likelihood_function.py / fit.py mirrors of the group package equivalents.
-- autolens_workspace_test multi_galaxy/jax_likelihood/ variant (only
-  model_fit + the relocated composition test landed).
+  BLOCKED from cloud sessions (2026-07-26): MAST is unreachable through the
+  session proxy (`Tunnel connection failed: 403`) — needs a
+  local/unrestricted-network session to download + prepare the frames.
+- Extra-galaxies / pixelization feature variants remain README cross-links
+  (the group/imaging feature scripts apply verbatim with the lens loop).
 
 Create the new `scripts/multi_galaxy/` package in @autolens_workspace — the first
 of the three above-galaxy-scale regimes (see the parent plan for the full design

draft/docs/autolens/split_lensing_regimes.md

@@ -307,13 +307,23 @@ regenerate the dataset alongside the convention change.
 1. `draft/docs/autolens/multi_galaxy_package.md` — new
    `autolens_workspace/scripts/multi_galaxy/` package (+ workspace_test
    mirror). Unblocks everything user-facing; do first.
+   CORE SHIPPED 2026-07-25; features/scaling_galaxies, fit.py and the
+   workspace_test jax_likelihood variant 2026-07-26. Remaining: real
+   J1011+0143 data swap-in (MAST unreachable from cloud sessions — needs a
+   local/unrestricted-network session).
 2. `draft/docs/workspaces/group_halo_explicit_choice.md` — group start_here
    gains all three tiers; new `features/group_halo/` halo-choice tutorial.
+   CORE SHIPPED: halo tutorial 2026-07-25; three-tier start_here (with an
+   image-derived scaling_galaxies.csv for the real Euclid dataset)
+   2026-07-26. Remaining: modeling/simulator halo-narrative threading.
 3. `draft/docs/workspaces/cluster_regime_narrative.md` — cluster narrative
    alignment + `features/extended_source/` follow-up example + conventions
    cross-check.
 4. `draft/docs/autolens/docs_three_regime_restructure.md` — PyAutoLens RTD
-   docs ladder rewrite (after 1, so links resolve).
+   docs ladder rewrite (after 1, so links resolve). SHIPPED 2026-07-26:
+   New User Guide ladder + flagships, overview_1 enumerations, api
+   mass/point/galaxy regime notes + CSV-catalogue section, model_cookbook
+   regime recipes.
 5. `draft/docs/workspaces/galaxy_scale_scaling_extra_features.md` —
    imaging/interferometer/point_source extra_galaxies + scaling_galaxies
    feature parity with regime caveats.

draft/docs/workspaces/group_halo_explicit_choice.md

@@ -8,7 +8,7 @@ Repos:
 Difficulty: large
 Autonomy: supervised
 Priority: high
-Status: in progress — signature tutorial landed on branch claude/pyautolens-doc-reorganization-w6a1l5 (2026-07-25)
+Status: in progress — signature tutorial landed 2026-07-25; three-tier start_here landed 2026-07-26 (branch claude/pyautolens-doc-reorganization-w6a1l5)
 Parent: draft/docs/autolens/split_lensing_regimes.md
 
 ## Landed (2026-07-25, this task branch)
@@ -22,11 +22,26 @@ Parent: draft/docs/autolens/split_lensing_regimes.md
   features README + regime-ladder README edits landed with the multi_galaxy
   package commit.
 
+## Landed (2026-07-26, this task branch)
+
+- `group/start_here.py` default model is now three-tier: main (2 galaxies —
+  the central galaxy plus the bright companion 0.36" away, promoted from
+  the extras catalogue after image inspection confirmed a distinct peak;
+  each MGE + Isothermal), extra (1, from `extra_galaxies_centres.json` —
+  MGE with a ±0.1" uniform light-centre prior + mass-centre-fixed bounded
+  `IsothermalSph`), scaling (5, mass-only untruncated `IsothermalSph` on a
+  shared `einstein_radius_ref` prior via
+  `einstein_radius_ref * luminosity_ratio**0.5`; the truncated-dPIE variant
+  stays pointed at `features/group_halo`). Scaling-galaxy
+  centres/luminosities for the real Euclid dataset were derived from the
+  image itself (gaussian-smoothed peak detection outside the main pair,
+  1"-aperture photometry, normalized to the brightest) and committed as
+  `dataset/group/.../scaling_galaxies.csv` with provenance documented in
+  the script prose. 24-free-parameter model validated end-to-end under
+  test mode.
+
 ## Remaining
 
-- `start_here.py` default model gains the extra_galaxies + scaling_galaxies
-  tiers (currently main_galaxies only) — needs scaling-galaxy
-  centres/luminosities prepared for the real Euclid dataset.
 - `modeling.py`/`simulator.py` halo-narrative threading; CSWA 19 as the
   possible future real-data flagship for this feature (public HST +
   published PyAutoLens model, arXiv:2504.11445).