analyze-results

Analyze machine learning experiment results into comparison tables and statistical summaries.

Updated May 29, 2026
One-click install
npx skills add https://github.com/Mang30/myskills --skill analyze-results-mang30
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: analyze-results
Source: https://github.com/Mang30/myskills/tree/main/skills/analyze-results
Command: npx skills add https://github.com/Mang30/myskills --skill analyze-results-mang30

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you interpret machine learning experiment outputs quickly by turning raw JSON/CSV result files into understandable comparisons, statistics, and actionable findings.

Core Features & Use Cases

  • Comparison tables with deltas vs baseline: Organizes results by independent variables and computes relative improvement for the main metric.
  • Statistical summaries across seeds and sweeps: Reports mean ± std when multiple seeds exist and identifies trends when parameters are swept.
  • Insight generation with next-step recommendations: Produces numbered observations, interpretations, implications, and proposed experiments to validate hypotheses.

Example use case: You ran a hyperparameter sweep for a model and produced several result files across different configurations; this Skill helps you identify the best configuration, quantify improvements versus baseline, and suggest the next experiment to test why it works.

Quick Start

Ask the Skill to analyze your experiment results by providing the path or description of the results directory that contains the JSON/CSV files.

Frequently Asked Questions about analyze-results

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

FAQPage Schema
How do I analyze machine learning experiment results from multiple JSON and CSV files?

To analyze machine learning experiment results, this Skill discovers raw JSON and CSV outputs in your project directories, then structures them into comparison tables and statistical summaries to quantify independent variable impacts.

How do I compare model performance against a baseline across hyperparameter sweeps?

Comparing model performance against a baseline involves structuring independent variables, computing relative metric deltas, and reporting mean ± std across multiple seeds to identify trends during hyperparameter sweeps.

What is the best way to generate actionable research insights from ML experiment outputs?

Generating actionable research insights from ML experiment outputs requires computing statistical summaries, formatting raw results into comparison tables, and producing numbered observations with proposed next experiments to validate hypotheses.

Can I use this to interpret machine learning results stored in figures and results folders?

Yes, you can interpret machine learning results by targeting common project folders like figures/ and results/, allowing the Skill to discover relevant mixed JSON and CSV output files for automated data interpretation.