Paper Figure

wanshuiyin/auto-claude-code-research-in-sleep

Automates publication-quality plots and LaTeX tables from experiment data. Handles the data-driven figures in a typical ML paper: training curves, bar charts, comparison tables, heatmaps, multi-panel grids. The scope is clear upfront: it won't generate architecture diagrams or qualitative result grids, those are manual. It reads your figure plan from PAPER_PLAN.md, creates individual Python scripts with consistent matplotlib styling (300 DPI, serif fonts, clean spines), runs them all, and outputs LaTeX snippets ready to paste. Supports Chinese commands like "画图" alongside English. The workflow is sensible: parse plan, set up shared style config, auto-select chart type based on data pattern, generate scripts, verify outputs, bundle LaTeX includes. Honest about covering roughly 60% of figures in a real paper, leaving hero figures and diagrams to you.

skills.sh installs: 657
GitHub stars: 18.9K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add wanshuiyin/auto-claude-code-research-in-sleep --skill paper-figure --agent claude-code

Installs into .claude/skills of the current project.

Files

Paper Figure: Publication-Quality Plots from Experiment Data

Before each reviewer slot, follow Selectable Reviewers: resolve invocation → .aris/reviewer.json → Claude Code default (Codex MCP) with reviewer_route.py through the canonical helper resolver, preserving this slot's tier, fresh contexts, source/image inputs and deterministic gates. Codex examples/pins below apply only to backend=codex; other routes use their own first-call/reply contract. Save the route and actual returned identity; unavailable selections are REVIEW_UNAVAILABLE. Classify acceptance from actual families/completion/scope. Forward invocation reviewer/model/effort overrides to reviewer-bearing children. Generation workers are unchanged.

Generate all figures and tables for a paper based on: $ARGUMENTS