← ClaudeAtlas

figrecipelisted

Reproducible matplotlib figures with mm-precision layouts, declarative specs, YAML recipes, multi-panel composition, and diagram generation. Use when creating publication-ready figures.
scitex-ai/claude-code-telegrammer · ★ 1 · AI & Automation · score 61
Install: claude install-skill scitex-ai/claude-code-telegrammer
# figrecipe ## Installation ```bash pip install figrecipe # Development: pip install -e /home/ywatanabe/proj/figrecipe ``` Reproducible, style-editable scientific figures via YAML recipes. Drop-in replacement for `plt.subplots()` that records all plotting calls for exact reproduction. ## Sub-skills * [quick-start](quick-start.md) — Core workflow: subplots, save, reproduce * [plot-types](plot-types.md) — All supported plot types with examples * [composition](composition.md) — Multi-panel figure composition (grid and mm-based) * [cropping](cropping.md) — Figure cropping, whitespace removal * [styles](styles.md) — Style presets, SCITEX/MATPLOTLIB, custom styles * [bundle](bundle.md) — ZIP bundle format (spec.json + data.csv + exports) * [diagram](diagram.md) — Box-and-arrow diagrams, Mermaid, Graphviz * [cli-reference](cli-reference.md) — All CLI commands * [mcp-tools](mcp-tools.md) — MCP tool reference for AI agents * [return-fig](return-fig.md) — Convention: plotting functions must return fig ## MCP Tools | Tool | Purpose | |------|---------| | `plt_plot` | Create figure from declarative spec dict | | `plt_reproduce` | Reproduce figure from YAML recipe | | `plt_compose` | Compose multi-panel figure | | `plt_crop` | Crop whitespace from figure image | | `plt_validate` | Validate recipe reproducibility | | `plt_extract_data` | Extract plotted data arrays | | `plt_info` | Get recipe metadata | | `plt_get_plot_types` | List all supported plot types | | `plt_list_styles` | L