| name | paper-figure |
| description | Generate publication-quality matplotlib or seaborn figures for an academic LaTeX paper. 11 EECS chart types (bar, line, heatmap, ROC, ablation, scaling). Output to figures/*.pdf. |
| argument-hint | <chart-type> <description> |
| paths | ["**/main.tex","**/figures/","**/scripts/*.py","**/.paper-config.yml"] |
| allowed-tools | Read Write Edit Bash(python *) Skill |
paper-figure
Generate a publication-quality figure for an academic LaTeX paper using
matplotlib or seaborn, save it as PDF in figures/, and write the matching
Python source in scripts/.
Project config loading (mandatory first step)
Same as paper-draft (see that file for the exact 5-step protocol). If no
.paper-config.yml is found in the cwd ancestor chain, stop with the
"No paper project found" error.
What this skill does
- Parse
$ARGUMENTS for <chart-type> and <description>.
- Recognized chart types:
bar, line, scatter, heatmap, roc,
ablation, scaling, confusion-matrix, box, violin, pareto.
- Delegate to the
visualization agent. The agent:
- Generates a Python script using matplotlib/seaborn with publication
defaults (300 DPI, colorblind-safe palette, proper rcParams)
- Saves the figure as
figures/<slug>.pdf
- Saves the script as
scripts/<slug>.py
- Returns the LaTeX include snippet for the user to paste
- Run the script with
python3 scripts/<slug>.py to actually produce the
PDF. Verify the output file exists and is non-empty.
Handoff schemas
None. This skill is self-contained.
Next steps
/paper-draft <section> to reference the new figure in the section text
/paper-compile to verify the figure renders in the PDF