metric-diagnostics

Name: metric-diagnostics
Availability: InStock
Author: OpenAI

Diagnose metric changes by validating drivers and generating calibrated explanations.

488|76|Updated Jun 2, 2026

Search Tags:#metrics #data-analytics #diagnose #reproducibility #driver-analysis

Ready to run

npx skills add https://github.com/openai/role-specific-plugins --skill metric-diagnostics

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: metric-diagnostics
Source: https://github.com/openai/role-specific-plugins/tree/main/plugins/data-analytics/skills/metric-diagnostics
Command: npx skills add https://github.com/openai/role-specific-plugins --skill metric-diagnostics

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Diagnose why a metric changes or differs from expectation by reproducing the metric, choosing the right comparison, validating likely drivers, and producing a calibrated explanation. Use when the user needs to understand what drove a metric movement, anomaly, gap, or discrepancy.

Core Features & Use Cases

Reproduce the metric behavior across contexts to verify consistency
Validate driving factors using multi-source evidence and live reads
Generate calibrated explanations suitable for dashboards and reports
Use cases include anomaly investigations, reconciliation, and trend explanations

Quick Start

Identify the diagnostic question and initiate a reproducible metric-and-driver analysis to surface the main driver.

Frequently Asked Questions about metric-diagnostics

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

FAQPage Schema

How do I diagnose why a metric changed or differed from expectation?▼

To diagnose metric changes, reproduce the metric across contexts, validate driving factors using multi-source evidence, and generate calibrated explanations for your reports.

What's the best way to identify driving factors behind a time-series anomaly?▼

Identify driving factors by performing driver decomposition on the time-series metric, verifying consistency through multi-source data reads, and validating likely drivers against live evidence.

How do I generate shareable reports for metric anomaly investigations?▼

Generate shareable reports for anomaly investigations by producing calibrated explanations and artifacts that summarize the diagnosed metric movements, gaps, and discrepancies.

When do I need to use driver decomposition for data reconciliation?▼

Use driver decomposition for data reconciliation when you need to understand what drove a metric movement, validate likely drivers across multi-source data, and calibrate the explanation.

Does this approach work for diagnosing gaps and discrepancies across analytics workflows?▼

Yes, diagnosing gaps and discrepancies is applicable across analytics workflows, reproducing metric behavior to verify consistency and validate driving factors using multi-source evidence.

Can I use metric diagnostics to validate driving factors with multi-source data verification?▼

Yes, you can validate driving factors by executing multi-source data verification, reproducing the metric behavior, and generating calibrated explanations for shareable dashboards.