cosmo-agent-skills

Deep-read an academic paper and produce a durable, critical research note — not a loose summary. Use when given a paper title, DOI, arXiv ID, publisher/PDF/GitHub URL, or local PDF, or when asked to read closely, summarize, extract contributions, analyze experiments, critique, compare to related work, reproduce results, or maintain a paper-reading repository index. Works across fields (astro/physics, ML, bio, security, …); examples lean astro-ph but the workflow is general.

Remote, slice-level access to FLAMINGO HDF5 files on COSMA via the `hdfstream` Python module — no need to download whole snapshots. Covers install, server connection, navigation, dataset slicing, multi-slice batched reads, and integration with swiftsimio. Use when reading any FLAMINGO data product remotely.

FLAMINGO past-lightcone products: virtual observers, particle lightcones, HEALPix shell maps, integrated_maps, and halo lightcones. Use when building mock surveys, stacking Compton-y or lensing maps, or cross-correlating with observations.

Review code changes against requirements and quality standards before merge. Use when dispatching a review subagent, after major features, before merging, or when the user asks for a code review. Requires reading the actual diff.

Style and structure for research manuscript text: IMRaD, hourglass narrative, formal tone, topic sentences, reproducible Methods, captions that tell the story. Use when drafting or editing titles, abstracts, sections, captions, or reports.

Iterative VLM review of compiled LaTeX PDFs for layout clashes (text vs figures, tables, long equations). Use after the first compile or any float/equation edit; loop fix-recompile-recheck until no clashes remain.

End-to-end research paper workflow: plan the narrative, draft IMRaD sections, produce figure assets, compile LaTeX to PDF, and debug layout by visually inspecting output (VLM). Use when starting, editing, or submitting a manuscript.

Numerical correctness check on saved result arrays (npz, HDF5, torch, etc.). Use after plot-check or when validating metrics without a new figure. Domain laws, bounds, and thresholds live in the project's CLAUDE.md, not here.

Visual (VLM) and array-completeness check for plots and figure files. Use proactively after savefig or export; before citing a figure in a paper or report. Axis styling and export settings → scientific-plotting; domain physics → physics-check.

Tracks which script and run produced each manuscript figure, plus VLM validator status. Use when adding figures, auditing stale results, or long agent runs.

Run plot-check then physics-check on a new figure and its arrays. Use proactively after every script that writes a plot or result file. Prefer plot-check or physics-check alone when only one layer applies.

Publication-quality plots: vector PDF output, PNG at dpi=300, label and tick font sizes readable at final column width. Use when generating or updating figures for papers, slides, or reports; or when plot text is too small or clipped.

Uses vision (VLM) to audit figures, tables, and layout in manuscript PDFs before submission. Primary debug path for publication quality, float clashes, and legibility. Use when creating plots, editing papers, recompiling PDFs, or when the user mentions visual QA or figure validation.

Audit and tidy a repo or working folder: fix layout drift, consolidate misplaced files, remove stale artifacts, align docs with disk. Use when organizing any project, scratch folder, or new repo — before release or after a refactor.

FLAMINGO SOAP halo catalogues and HBT-HERONS subhalo outputs: directory layout, halo definitions (bound, spheres, projected apertures, spherical overdensity), and cross-simulation matching. Use when selecting clusters or galaxies from released catalogues.

FLAMINGO precomputed matter and component power spectra (ASCII P(k), 123 redshift indices), units without h, and the FlamingoBaryonResponseEmulator. Use when loading FLAMINGO P(k), comparing to theory, or emulating baryonic suppression without full hydro.

FLAMINGO simulation suite inventory: naming (L box, m resolution, variations, _DMO), fiducial and variation runs, directory layout under /FLAMINGO/. Use when choosing a run, locating data on the COSMA portal, or citing the simulation.

FLAMINGO SWIFT particle snapshots: virtual HDF5 files, particle types, 78/79 output redshifts, downsampled and reduced variants. Use when reading gas, DM, stars, BHs, or neutrinos from a FLAMINGO run.

Test-driven workflow: write failing tests first, verify red/green, cover edge cases, and keep the full suite green before merge. Use when implementing features, fixing bugs, or refactoring — before and alongside production code.

Guidelines for implementing code changes: think first, keep diffs minimal and simple, surface assumptions, and verify with tests including edge cases. Use when writing new code, fixing bugs, or refactoring — before and during implementation.