analyze-results

Generate analysis-report.json from the latest result group for drift review.

1|Updated Mar 23, 2026

Search Tags:#workflow #analysis #evidence #drift #analyze #results

Ready to run

npx skills add https://github.com/LinkDry/Link-Research --skill analyze-results-linkdry

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: analyze-results
Source: https://github.com/LinkDry/Link-Research/tree/main/skills/analyze-results
Command: npx skills add https://github.com/LinkDry/Link-Research --skill analyze-results-linkdry

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Interpret the latest result group into a structured analysis artifact to support drift and judge reviews.

Core Features & Use Cases

Interpret the latest result group into analysis-report.json, preserving evidence while deferring verdicts.
Update experiment state and project steering state to reflect the new analysis, enabling drift-detector and judge skills to proceed.

Quick Start

After a new result group is recorded, run analyze-results to generate analysis-report.json and prepare for drift review.

Frequently Asked Questions about analyze-results

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

FAQPage Schema

How do I analyze experiment results to prepare for drift review?▼

To analyze experiment results for drift review, run the analyze-results process to transform the latest result group into a structured analysis artifact. This generates an analysis-report.json file that preserves evidence while deferring verdicts, enabling subsequent drift and judge reviews.

What is an analysis artifact and how does it support workflow evidence?▼

An analysis artifact is a structured report that interprets experiment results while preserving evidence. It supports workflow evidence by recording a conservative interpretation of the latest result group, enabling drift-detector and judge skills to proceed without losing raw data context.

How do I generate an analysis report from a new result group?▼

Generate an analysis report by running the analyze-results task after a new result group is recorded. The process checks for an active line with a latest_result_ref and no existing analysis artifact, then writes analysis-report.json to the workspace results directory.

Does the analysis process automatically apply drift and judge decisions?▼

The analysis process does not automatically apply drift and judge decisions. It strictly interprets results into a structured analysis artifact, updating experiment-memory.md and STATE.md while explicitly leaving drift and judge verdicts to subsequent skills in the workflow.

When do I need to run structured analysis on my experiment results?▼

Run structured analysis on experiment results when a new latest_result_ref exists for an active line and no analysis artifact currently exists. This ensures conservative interpretation occurs before drift review, maintaining workflow integrity and preparing state files for downstream processing.

What state files are updated when interpreting result groups?▼

When interpreting result groups into an analysis artifact, the process updates experiment-memory.md and STATE.md to reflect the new analysis. These updates enable downstream drift-detector and judge skills to access the current analysis state and proceed with their respective reviews.