Skip to content

/figure-comparison

Build a self-contained HTML report (.lightcone/comparison.html) that places paper reference artifacts on the left and reproduced artifacts on the right, with red flags wherever a counterpart is missing. Images are base64-embedded so the HTML is portable. Run from a project folder containing astra.yaml.

Source: claude/lightcone/skills/figure-comparison/SKILL.md.

Argument hint: [path to paper reference dir, e.g. work/reference/].

Allowed tools

Read, Write, Glob, Grep,
Bash(ls:*), Bash(wc:*), Bash(grep:*), Bash(find:*), Bash(file:*),
Bash(python3:*), Bash(python:*), Bash(base64:*),
AskUserQuestion, Agent

Read-only over the build artifacts. The skill never invokes the pipeline itself — if results/<universe>/ is empty, it tells the user to run lc run first and stops.

Setup

  1. Confirm project root. Reads astra.yaml in the cwd. If missing, asks the user to cd to the ASTRA project.
  2. Confirm results exist. Default universe is baseline, unless comparison-report.yaml names another universe or the user supplied one. Checks ls results/<universe>/.
  3. Locate the paper reference substrate. In order: a path passed as an argument, then work/reference/ from lc-from-paper's layout — index.json (the canonical structural index of figures and tables, written by /paper-extraction on both paths), plus source/ for arXiv TeX or document.md for the Docling fallback, plus the extracted figures/ and tables/. Legacy locations are tried only after lc-from-paper paths fail. A stray metadata.json from the Docling path is ignored — its content is already folded into index.json.

Scope resolution

The skill picks its target set in priority order:

  1. comparison-report.yaml — the highest-priority scope when lc-from-paper has run COMPARE. Records exactly what to compare, including type, priority, paper/reproduced values, file paths, and match status.
  2. targets/targets.md — the SPECIFY-phase scope ledger, used when COMPARE hasn't run yet.
  3. Default paper-driven flow — when neither scope file exists, builds a best-effort report from astra.yaml's description and findings: plus work/reference/.

Resolving the reproduced side

Every lc run output is a directory — results/<universe>/<output_id>/ containing the artifact file(s) plus .lightcone-manifest.json. The skill globs inside that directory for the first suitable type-specific extension (images, tables, values), always ignoring .lightcone-manifest.json and .snakemake_timestamp. An explicit reproduced_file from the scope file wins; filename-stem similarity across results/<universe>/*/ is the last resort, and an unmatched target renders as a red NOT PRODUCED panel rather than being guessed at.

Output

A single .lightcone/comparison.html with paper artifacts on the left and reproduced artifacts on the right. Helper scripts and intermediate manifests also live under .lightcone/ so they don't pollute the baseline results. The skill never creates .lightcone/ — it exists in any lc init-ed project, and its absence means the project isn't initialized.

The HTML embeds figure images as base64 — paste it into email, drop it on a shared drive, or send it through Slack without breaking links.

When to invoke

  • From /lc-from-paper's REVIEW close-out (mandatory).
  • Standalone, any time after lc run succeeds, to see how the reproduction stacks up against the paper.

Hard rules

  • Read-only over build artifacts. Never run the pipeline; if outputs are missing, stop and ask the user to build first.
  • Don't compare directly against a whole PDF. When only work/reference/paper.pdf exists, ask the user to run /paper-extraction first (in lc-from-paper projects this happens during ORIENT).
  • Preserve scope ordering. comparison-report.yaml wins over targets/targets.md wins over the default flow.