dashboard.md

PyAutoMind backlog dashboard

133 filed prompts in the backlog · 7 already dispatched to issues (active/). Backlog view only — organism health lives with the Heart (/health), not here.

Work-type	Prompts
bug	39
feature	25
maintenance	21
research	18
docs	17
refactor	5
test	3
triage	3
experiment	1
release	1

bug (39)

Prompt	Target	Difficulty	Autonomy	Priority
PROBE: is Adapt's 4th-power coefficient dependence (double square) intentional?	autoarray	medium	supervised	normal
ConstantZeroth regularization is broken twice over — dead code presenting	autoarray	small	supervised	normal
pixel_scales given as an int (or np.float64) is never widened	autoarray	small	supervised	medium
PyNUFFT dev extra is incompatible with current SciPy on Python	autoarray	small	supervised	normal
EP: cure the hierarchical parent-scale collapse basin (and make F10	autofit	medium	supervised	high
autofit.plot functions accept **kwargs and silently discard them	autofit	small	supervised	normal
TEST_MODE bypass crashes on ordered-parameter assertion ties	autofit	small	supervised	normal
python_matrix smoke fails: autofit_workspace searches/mle.py needs optax not in smoke	autofit_workspace	small	safe	normal
NFWTruncatedSph.potential_2d_from: MGE potential fails grad(psi)=alpha self-consistency	autogalaxy	too-large	supervised	high
interferometer Delaunay pixelization — non-PD FitException in test-mode bypass	autolens	medium	supervised	normal
interferometer/start_here.py OOM in nightly release-validation integrate leg	autolens	-	-	-
JAX point-source smoke sentinel: point.py returns -1e99 instead of -83.38	autolens	medium	supervised	normal
JIT cache not hit in modeling_visualization delaunay/rectangular scripts	autolens	medium	supervised	normal
Investigate eager FitImaging.figure_of_merit vs JIT/step-by-step divergence in rectangular pixelization	autolens	too-large	supervised	high
point.py JAX-vmap parity assert is non-deterministic under the smoke env	autolens	small	supervised	normal
hpc/sync first-push race — parallel rsyncs before remote base dir	autolens_assistant	small	safe	normal
Scripts derive geometry from a hardcoded pixel_scale while the dataset	autolens_workspace	small	supervised	normal
jax_grad scripts fail assertions locally that PASS in CI	autolens_workspace_test	medium	supervised	medium
add_notebook_quotes mistakes a code string literal's closing delimiter for a	hands	small	safe	low
Fix Autofit release sampler and database regressions	health_fixes	too-large	supervised	high
Fix release JAX runtime compatibility and likelihood parity	health_fixes	too-large	supervised	high
Fix JIT quick-update visualization output regressions	health_fixes	too-large	supervised	high
Fix release-profile numerical inversion failures	health_fixes	too-large	supervised	high
Resolve release-profile timeout scripts deliberately	health_fixes	too-large	supervised	normal
Fix release result/sample parameter-path regressions	health_fixes	too-large	supervised	high
Audit HowTo tutorials for missing setup_notebook() line	howto	small	safe	normal
HowToGalaxy small API drifts: ellipse kwargs + plot_grid_lines (parked NEEDS_FIX)	howtogalaxy	small	supervised	normal
@PyAutoFit Add property-based correctness tests for every Prior subclass	priors	large	supervised	normal
@PyAutoFit TransformedMessage reversal convention is undocumented foot-gun	priors	large	supervised	normal
@PyAutoFit Refactor: each density should live in one place, not	priors	too-large	supervised	normal
@PyAutoFit Refactor: collapse the Prior / Message two-layer hierarchy	priors	too-large	supervised	normal
@PyAutoFit Refactor: replace hand-rolled AbstractDensityTransform with tfp.bijectors / numpyro.distributions.transforms	priors	too-large	supervised	normal
Priors & Messages cleanup — tracker	priors	too-large	supervised	normal
Release does not sync version stamps and workspace pins back	pyautobuild	medium	supervised	high
generate.py deletes notebooks/ before rejecting an unknown project	pyautohands	small	supervised	normal
pre_build stages untracked files, publishing uncommitted human work	pyautohands	small	supervised	high
Heart script_timing baselines are orphaned by path moves and filled	pyautoheart	small	supervised	medium
Tenant firewall: release_run.py carries an unlisted 'PyAutoLabs' instance fact	pyautoheart	small	safe	normal
aplt.Output stale-API drift in the remaining workspace repos	workspaces	small	supervised	normal

feature (25)

Prompt	Target	Difficulty	Autonomy	Priority
Claude Development Prompt: Arcsecond Tick Label Decimal Placement	autoarray	large	supervised	normal
Can create a list of InversionMatrix objects for each dataset	autoarray	medium	supervised	normal
Follow-up to rectangular_adapt_cdf.md (issue #322) and Path A	autoarray	too-large	supervised	normal
EP analytic updates — implement the four planned work packages	autofit	large	supervised	normal
The project @z_projects/ic50_workspace is our IC50 use case which we	autofit	medium	safe	normal
Give PyAutoFit searches a seed — today no search can	autofit	medium	supervised	medium
Remote-MCP deployment tiers (2 + 3) for the results-inspector server	autofit_assistant	large	human-required	normal
dPIE: optional central-dispersion (sigma_0) parameterization	autogalaxy	small	supervised	low
PIEMass.potential_2d_from: implement the missing lensing potential	autogalaxy	too-large	supervised	normal
autolens_jax_joss benchmark repo + real-data start_here pairing	autolens_jax_joss	too-large	supervised	normal
Profile and speed up JAX likelihood-function compile times (all use	autolens_profiling	large	supervised	high
Search settings-estimation + profiling infrastructure (n_starts / batch_size / n_batch)	autolens_profiling	large	supervised	normal
Tune cluster-scale JOSS benchmarks toward their 5-minute targets	autolens_workspace	medium	supervised	normal
Adopt oversampled PSFs in the start-here dataset chain (option a)	autolens_workspace	large	supervised	normal
Scheduled runs — overnight queue passes with a morning report	autonomy	medium	supervised	low
Context: PyAutoLens issue #542 follow-up (Gap 1, deferred during the	jax_substructure	too-large	supervised	normal
Context: PyAutoLens issue #542 follow-up (Gap 2, deferred during the	jax_substructure	too-large	supervised	normal
Give the Profiling Agent a compile-time axis — the arc	profiling	large	supervised	high
Token-light wiki index over the complete/ archive	pyautomind	medium	supervised	normal
Make draft/ staleness detectable — intake reconcile measured, and the	pyautomind	medium	supervised	high
LACosmic per-frame CR masking option + decouple PSF-star pass from	pyautoreduce	medium	supervised	high
Gallery runner: add visualization_upper + decide the modeling_visualization_jit tier	workspaces	small	supervised	low
The imaging features/advanced/los_halos example needs improving and padding out before	workspaces	medium	safe	normal
The imaging features/advanced/subhalo/sensitivity example needs improving and padding out before	workspaces	medium	safe	normal
Once https://github.com/PyAutoLabs/PyAutoLens/issues/480 is fixed (PointSolver	workspaces	too-large	supervised	normal

maintenance (21)

Prompt	Target	Difficulty	Autonomy	Priority
dataset/imaging/jwst_lw is untracked because the gitignore was never extended for	autolens_profiling	small	supervised	low
autolens_profiling is now a mature project, with a good separation	autolens_profiling	large	supervised	normal
cosmos_web_ring stores boolean masks as float64, wasting ~3.4 MB of	autolens_workspace	small	supervised	low
LaTeX in non-raw docstrings emits SyntaxWarning: invalid escape sequence	autolens_workspace	small	supervised	low
autolens_workspace_developer rectangular experiments — Gut stash + rename	autolens_workspace_developer	small	supervised	normal
autolens_workspace_developer: broad stale-API rot (56 symbols, no CI)	autolens_workspace_developer	medium	supervised	normal
Auto-request GitHub Copilot code review on every PR, org-wide	ci	large	supervised	normal
run_smoke.py: three runner variants across 10 repos, no sync mechanism	ci	medium	supervised	normal
Dependency-cap refresh 2026-08: safe bumps, astropy 8 decision, two dead	libraries	medium	supervised	normal
PyAutoNerves committed version stamp behind sibling consensus	libraries	small	safe	low
PyAutoFit CLI-noise batch: unclosed search.log handler + four small warning	pyautofit	small	safe	normal
PyAutoMemory canonical-key TODO sweep	pyautomemory	medium	supervised	normal
Single-source the "Never rewrite history" policy as a generated AGENTS.md	pyautomind	medium	supervised	normal
Silence the three autonerves-rooted CLI-noise sources (fits leak, pytest collection,	pyautonerves	small	safe	normal
Capped smoke datasets were committed as if they were real	workspaces	medium	supervised	normal
Mirror drifted library config keys into the workspace configs	workspaces	small	supervised	normal
Raw-string the LaTeX docstrings emitting SyntaxWarnings (HowToFit + HowToLens)	workspaces	small	safe	low
Regenerate setup_notebook-drifted notebooks in autogalaxy/autofit/HowToFit workspaces	workspaces	small	supervised	low
autolens_workspace	workspaces	too-large	supervised	normal
Refresh the stale .script_sizes.json snapshot in @autolens_workspace	workspaces	small	safe	low
Un-park imaging/features/scaling_relation/slam once PyAutoArray#431 merges	workspaces	small	supervised	normal

research (18)

Prompt	Target	Difficulty	Autonomy	Priority
Delaunay-family JAX modules never hit the persistent compilation cache	autoarray	medium	supervised	medium
PyAutoArray Delaunay interpolator's pure_callback vs vmap — minor efficiency follow-up	autoarray	too-large	supervised	low
Deep research: Can we speed up Delaunay in PyAutoArray?	autoarray	too-large	supervised	high
Kernel-CDF bandwidth defaults — config-dependent quality, investigate adaptivity	autoarray	medium	supervised	normal
Use readthedocs or migrate to GitHub docs	autobuild	small	supervised	normal
Census of priors and messages — confirmed bugs + redesign	autofit	too-large	supervised	high
Quick-update plotting cost — minutes per update, and it is	autolens	medium	supervised	medium
Cluster-scale gradient-search benchmark (Prodigy vs Nautilus, point-source)	autolens_profiling	-	-	-
Multi-band compile census completion — A100/multi-core + hetero GPU rows	autolens_profiling	small	supervised	low
We have lots of examples which profile how long JAX	autolens_workspace_developer	medium	supervised	normal
Expectation Propagation Scale-Up — Scoping	graphical_ep	too-large	supervised	high
Graphical Model Scale-Up — Scoping	graphical_ep	too-large	supervised	high
slope_hierarchy: methods write-up (NUTS headline, EP cautionary)	graphical_ep	medium	supervised	normal
slope_hierarchy: scale the hierarchical slope recovery to N=25–50	graphical_ep	medium	supervised	normal
Adopt Python 3.12 as the PyAuto ecosystem minimum	libraries	-	-	-
Checkerboard PSF-mismatch residual diagnostic — research + document + ingest	pyautomemory	medium	supervised	normal
Re-baseline the slacs0008 acceptance parity after the HAP-dedupe fix	pyautoreduce	small	supervised	normal
Chase the ~6% flux scale between PyAutoReduce and legacy SLACS	pyautoreduce	medium	supervised	low

docs (17)

Prompt	Target	Difficulty	Autonomy	Priority
Rewrite PyAutoCTI docs/api — 55 of 89 autosummary entries are	autocti	medium	supervised	normal
PyAutoLens RTD docs: three-regime restructure (multi_galaxy / group / cluster)	autolens	medium	supervised	high
multi_galaxy package: new regime package in autolens_workspace	autolens	large	supervised	high
Split lensing regimes: multi_galaxy / group / cluster (epic plan)	autolens	too-large	supervised	high
Regenerate autolens_workspace markdown/ so the MGE pages show sigma_min	autolens_workspace	small	supervised	normal
Phase 2 — drop the hand-written quick-update sentence from the	autolens_workspace	-	-	-
HowToLens ch4 tutorial 3: mask overlay is never actually drawn	howtolens	small	supervised	low
Markdown renderings batch 2a — leftovers (ellipse/modeling + PNG size)	pyautobuild	small	safe	low
add-vincken-2026-wiki-and-cite-in-euclid	workspaces	small	safe	normal
Assistants: regime-aware routing for multi_galaxy / group / cluster (follow-up)	workspaces	medium	supervised	low
Cluster package: point-source-default narrative + extended-source follow-up feature	workspaces	medium	supervised	high
extra_galaxies feature parity: point_source + multi_galaxy (both workspaces)	workspaces	medium	supervised	normal
plot coverage — follow-ups deferred from plot-coverage-gaps	workspaces	-	-	-
Advanced workspace guide: Preloads (PyAutoArray)	workspaces	too-large	supervised	high
Propagate the shear_galaxy-at-(0,0) idiom to group/ and cluster/	workspaces	small	supervised	normal
Rectangular mesh Enzi citation — user-workspace pixelization examples	workspaces	small	supervised	normal
Phase 2: Make workspace READMEs assistant-first	workspaces	-	-	-

refactor (5)

Prompt	Target	Difficulty	Autonomy	Priority
Vendor bessel_kve into autoarray and drop the tensorflow-probability dependency	autoarray	large	supervised	medium
Split Fitness.batch_size into lh_batch_size and latent_batch_size	autofit	small	supervised	normal
einstein_radius_jit_from: replace static init_guess with a JAX-native seed finder	autogalaxy	too-large	supervised	high
Slow imports: autolens 4.3s, autogalaxy 3.4s (hygiene perf tier, >3s	libraries	medium	supervised	normal
Remove the dead EDEN packaging tooling from PyAutoFit	pyautofit	medium	supervised	normal

test (3)

Prompt	Target	Difficulty	Autonomy	Priority
Re-baseline the MGE imaging JIT profiling regression value	autolens_workspace_developer	too-large	supervised	high
Restore absolute NumPy likelihood regression baselines in the _workspace_test	workspaces	too-large	supervised	high
The new workspace smoke-test GitHub Actions (added via feature/smoke-test-ci) surfaced	workspaces	too-large	supervised	normal

triage (3)

Prompt	Target	Difficulty	Autonomy	Priority
Triage: Convolver "No blurring_image provided" warning in canonical workspace scripts	-	small	supervised	normal
<!-- TRIAGE: needs manual review before routing	-	medium	safe	high
Nightly release has been blocked 8 nights running — triage	-	medium	supervised	normal

experiment (1)

Prompt	Target	Difficulty	Autonomy	Priority
Tune the JAX multi-start optimizers into a standard option (MGE	autolens_profiling	large	supervised	high

release (1)

Prompt	Target	Difficulty	Autonomy	Priority
CTI release-train wiring — first modern autocti release	autocti	medium	human-required	normal

Hygiene

6 prompt(s) without a metadata header — they show - above. Re-home or re-run intake on them when touched.

Headerless prompts

• draft/bug/autolens/interferometer_release_leg_oom.md
• draft/docs/autolens_workspace/sampler_cli_output_workspace_sweep.md
• draft/docs/workspaces/plot_coverage_followups.md
• draft/docs/workspaces/unify_ai_assistant_workspace_readmes.md
• draft/research/autolens_profiling/cluster_gradient_search_benchmark.md
• draft/research/libraries/python_312_minimum.md