investigate-metric

Name: investigate-metric
Availability: InStock
Author: PostHog

Diagnose product metric changes via PostHog breakdowns, funnels, and SQL queries.

75|14|Updated Feb 18, 2026

Search Tags:#analytics #investigate #root-cause #posthog #cohort #metric-change #breakdowns

Ready to run

npx skills add https://github.com/PostHog/ai-plugin --skill investigate-metric

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: investigate-metric
Source: https://github.com/PostHog/ai-plugin/tree/main/skills/investigate-metric
Command: npx skills add https://github.com/PostHog/ai-plugin --skill investigate-metric

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Diagnose why a product metric changed by orchestrating breakdowns, actors, paths, lifecycle, retention, and annotations to surface root causes.

Core Features & Use Cases

Orchestrates multi-step root-cause analyses across trends, funnels, retention, and lifecycle metrics.
Enables cross-segment comparisons, actor drilldowns, and lifecycle partitioning to isolate causes.
Automates variance checks against priors using helper scripts for prior-period comparison and breakdown attribution.

Quick Start

Describe the metric change and time window, and let the skill orchestrate breakdowns, comparisons, and root-cause analysis.

Frequently Asked Questions about investigate-metric

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

FAQPage Schema

How do I find the root cause of a metric change in PostHog?▼

Diagnosing a metric change involves orchestrating breakdowns, actor drilldowns, and lifecycle partitioning across trends and funnels. This skill automates variance checks against priors to isolate root causes.

What is the best way to investigate anomalies in my product analytics dashboards?▼

Investigating anomalies in product analytics dashboards is best handled by applying multi-step root-cause analysis. This includes cross-segment comparisons, path analysis, and automated breakdown attribution to surface why the metric shifted.

Can I use SQL queries to diagnose why a PostHog trend metric dropped?▼

Yes, you can diagnose why a PostHog trend metric dropped by executing SQL queries and reading data schemas. The skill uses these alongside trend and funnel queries to perform variance checks and isolate the drop.

How do I perform a cohort breakdown attribution for metric variance?▼

Performing cohort breakdown attribution for metric variance requires comparing current segments against prior periods. Helper scripts automate this variance check and attribute the metric change to specific cohort breakdowns.

Does root-cause analysis work with pasted queries and saved insights?▼

Root-cause analysis works directly with saved insights, dashboards, and pasted queries. You describe the metric change and time window, and the analysis orchestrates breakdowns and comparisons across those inputs.

When should I use lifecycle partitioning to investigate a metric change?▼

You should use lifecycle partitioning to investigate a metric change when you need to isolate causes based on user retention and lifecycle stages. It helps cross-segment comparisons by categorizing actors into new, returning, and resurrected groups.