self-improving-science

Consolidate research learnings and experiment issues into centralized Markdown logs.

Updated Apr 13, 2026

Search Tags:#methodology #governance #learnings #reproducibility #experiments #model-card #data-leakage

Ready to run

npx skills add https://github.com/jose-compu/self-improving-skills --skill self-improving-science

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: self-improving-science
Source: https://github.com/jose-compu/self-improving-skills/tree/main/self-improving-science
Command: npx skills add https://github.com/jose-compu/self-improving-skills --skill self-improving-science

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Centralizes capture of learnings, experiment issues, and methodology corrections into structured Markdown logs to improve reproducibility and knowledge sharing across scientific workflows.

Core Features & Use Cases

Log learnings to LEARNINGS.md, track experiment issues in EXPERIMENT_ISSUES.md, and file feature requests in FEATURE_REQUESTS.md
Promote broadly applicable insights to experiments checklists, model cards, data governance docs, or tooling guidance
Link learnings to related files, datasets, and models to build a traceable knowledge base for audits and future experiments

Quick Start

Create the .learnings directory and initial log files, then begin recording your first learning entry.

Frequently Asked Questions about self-improving-science

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I track machine learning experiment issues and data leakage for reproducibility?▼

Log experiment issues and data leakage by recording methodological corrections in centralized Markdown files like EXPERIMENT_ISSUES.md. This structured logging builds a traceable knowledge base linking datasets and models to ensure reproducibility across data science workflows.

What is the best way to document research learnings and hypothesis revisions in ML projects?▼

Document research learnings and hypothesis revisions by maintaining centralized Markdown logs within a dedicated directory. Promoting these broadly applicable insights into model cards and data governance docs ensures reliable improvements and knowledge sharing across future experiments.

How do I set up structured logs for reproducibility and statistical mistakes in scientific workflows?▼

Set up reproducibility logs for statistical mistakes by creating a dedicated directory with specific Markdown templates for learnings, experiment issues, and feature requests. This structured environment captures methodological corrections and promotes findings into experiments checklists or governance docs.

Can I use Markdown logs to promote findings to model cards and data governance docs?▼

Yes, you can promote findings from Markdown logs directly into model cards and data governance docs. This requires dedicated tooling to elevate broadly applicable insights captured during experiments into formal governance records for reliable improvements.

Does capturing research learnings require a specific directory structure for experiments?▼

Capturing research learnings requires a dedicated directory containing specific Markdown templates for learnings, experiment issues, and feature requests. This structured setup ensures methodological corrections, drift observations, and data leakage findings are properly centralized for audits.