Which PostHog tools are required for this skill?

This skill utilizes experiment-get, execute-sql (for HogQL queries), and query-session-recordings-list to aggregate data.

Can it handle multivariate experiments with more than two groups?

Yes, it identifies all variants defined in the underlying feature flag and can construct filters for each group individually.

Does it exclude internal test accounts from the analysis?

Yes, the workflow is configured to set filter_test_accounts to true, ensuring your analysis is based on real user data.

How does the skill distinguish between experiment variants?

The skill uses the $feature/ event property to filter recordings specifically for users exposed to each distinct variant key.

Do I need an active experiment to use this skill?

Yes, the experiment must be launched and have recorded user traffic to provide meaningful session replays for analysis.

PostHog Experiment Replay Analyzer

Name: PostHog Experiment Replay Analyzer
Author: PostHog

This skill enables Claude to bridge the gap between quantitative experiment metrics and qualitative user behavior. By automatically filtering session recordings based on experiment variants (control vs. test groups), it helps teams identify exactly why certain variations perform differently. It streamlines the complex workflow of fetching experiment metadata, identifying feature flag variants from source-of-truth configurations, and performing deep dives into user interactions, console errors, and navigation paths to provide a comprehensive understanding of experimental impact.

Key Features

01Identification of qualitative friction points in specific user journeys

02Automated experiment metadata and feature flag variant retrieval

03Dynamic session recording filtering for control and test groups

04Direct integration with PostHog HogQL and recording list tools

05Comparative analysis of session duration, activity levels, and error rates

0659 GitHub stars

Use Cases

01Comparing user navigation patterns between two different checkout flow designs

02Supplementing statistical experiment results with visual evidence of user frustration or success

03Investigating why a specific experiment variant is triggering unexpected console errors

Tieline — product intent grounded in code

PostHog Experiment Replay Analyzer

Key Features

Use Cases

Energy Procurement Specialist

ECC Tools Cost Audit

Energy Procurement & Strategy

Key Features

Use Cases