mcpField
F
figures4papers-playbook

figures4papers-playbook

github.com/l-yifanData & databases538 installsUpdated 3w ago
Repository

About figures4papers-playbook

Figures4papers Playbook is a skill for locating a relevant plotting example and adapting it into a project-specific script. It uses chart type and domain to guide template selection, with an optional handoff to scientific-figure-pro for style polishing.

Key features

  1. 01Locate examples by keyword and chart type
  2. 02Use domain and folder context to select templates
  3. 03Adapt data, labels, and output paths
  4. 04Provide run commands and expected files

Use cases

  1. 01Create a grouped-bar benchmark comparison
  2. 02Find a plotting template for a research domain
  3. 03Adapt an example script to new data
  4. 04Prepare PNG and PDF figure outputs

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/l-yifan/skills --skill figures4papers-playbook

After installing, type/figures4papers-playbookin Claude Code to confirm it loaded.

Bundled filesreferences/example_index.mdscripts/example_locator.py

Details

From the project README

Figures4papers Playbook

Use this skill to find the closest figures4papers example and convert it into a user-specific plotting script quickly.

Workflow

  1. Parse user intent into chart type and domain (for example: grouped bar + benchmark comparison).
  2. Run scripts/example_locator.py to shortlist candidate scripts.
  3. Read references/example_index.md for folder-level context and naming patterns.
  4. Choose one base script and adapt data, labels, and file outputs.
  5. If style consistency is required, co-use scientific-figure-pro for final polishing.

Locator Script

Use keyword and chart filters:

python skills/figures4papers-playbook/scripts/example_locator.py --keyword bar --chart-type grouped_bar

Output includes:

  • Folder and domain.
  • Matching script paths.
  • Recommended run commands.
  • Notes about the intended figure purpose.

Selection Rules

  • Prefer exact chart-type match first.
  • Then prefer domain match (biomed, LLM eval, vision survey, etc.).
  • If no exact match exists, choose the simplest script with nearest visual grammar.
  • Keep output paths under a dedicated figures/ directory in the target project.

References

  • references/example_index.md: curated index of available figure folders and scripts.

Handoff Pattern

After locating a base script, provide:

  1. Selected template path.
  2. What to replace (data arrays, labels, colors, limits).
  3. Command to run.
  4. Expected output files (.png and .pdf).

Comments

Markdown supported