Experiment Code

This generates and refines ML experiment code for research papers, handling the full cycle from initial implementation to debugging and plotting. It enforces a specific structure with separate run directories, JSON logging, and mandatory visualization outputs. The iterative improvement loop is interesting: it reads results, reflects on what worked, makes targeted edits, and keeps the best variant. Works with PyTorch or scikit-learn, explicitly avoids placeholder code, and includes safeguards like verifying your accuracy isn't accidentally zero. Best for when you're implementing experiments and need something more structured than ad-hoc scripting but don't want to set up a full MLOps pipeline.

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 experiment-code --agent claude-code

Installs into .claude/skills of the current project.

Files

Experiment Code

Generate and iteratively improve ML experiment code for research papers.

Input

$0 — Task: generate, improve, debug, plot
$1 — Research plan, idea description, or error message

References

Experiment prompts and patterns: ~/.claude/skills/experiment-code/references/experiment-prompts.md
Code patterns (error handling, repair, hill-climbing): ~/.claude/skills/experiment-code/references/code-patterns.md

Action: `generate`

Generate initial experiment code following this structure:

Plan experiments first — List all runs needed (hyperparameter sweeps, ablations, baselines)
Write self-contained code — All code in project directory, no external imports from reference repos
Include proper logging — Save results to JSON, print intermediate metrics
Generate figures — At minimum Figure_1.png and Figure_2.png

Mandatory Structure

project/
├── experiment.py      # Main experiment script
├── plot.py            # Visualization script
├── notes.txt          # Experiment descriptions and results
├── run_1/             # Results from run 1
│   └── final_info.json
├── run_2/
└── ...

SKILL.md on GitHub