Paper To Code

Takes a research paper PDF or text and generates a complete, runnable code repository through a three-stage pipeline: planning (UML diagrams, dependency graphs), analysis (per-file logic mapping), and coding (dependency-ordered generation). The workflow is methodical, maybe overly so, generating everything from config.yaml to reproduce.sh scripts while strictly following the paper's methodology without inventing improvements. Useful when you need to replicate published results rather than just understand the approach. The constraint to generate code in dependency order and never fabricate hyperparameters keeps it honest, though you'll still need to debug if the paper glosses over implementation details. Best for papers with clear experimental setups rather than purely theoretical work.

skills.sh installs: 1.6K
GitHub stars: 459
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add lingzhi227/agent-research-skills --skill paper-to-code --agent claude-code

Installs into .claude/skills of the current project.

Files

Paper to Code

Convert a research paper into a complete, runnable code repository.

Input

$0 — Paper PDF path, paper text, or paper URL

References

Paper2Code prompts (planning, analysis, coding stages): ~/.claude/skills/paper-to-code/references/paper-to-code-prompts.md

Workflow (from Paper2Code)

Stage 1: Planning

Four-turn conversation to create a comprehensive plan:

Overall Plan: Extract methodology, experiments, datasets, hyperparameters, evaluation metrics
Architecture Design: Generate file list, Mermaid classDiagram, sequenceDiagram
Task Breakdown: Logic analysis per file, dependency-ordered task list, required packages
Configuration: Extract training details into config.yaml