What problem does it solve?
Session replay silently breaks in two ways: recordings stop being captured while site traffic continues (a capture cliff that permanently loses data), and user friction like rage clicks or errors-after-interaction concentrates on specific pages without anyone noticing. This Skill continuously monitors PostHog session replay data to catch both failure modes and files corroborated, actionable reports.
Core Features & Use Cases
- Capture cliff detection: Compares daily recording volume against site traffic to spot unexplained drops, then diagnoses the cause via Team config activity logs and SDK health properties like
$replay_sample_rate.
- Friction concentration analysis: Finds URLs and elements where rage clicks, dead clicks, or errors-after-click spike above their own historical baseline, corroborated with heatmaps and watchable example recordings.
- Replay vision scanner monitoring: Aggregates LLM scanner observations across sessions and detects watch gaps where scanners silently stop producing observations.
- Use Case: A PostHog user's checkout page suddenly shows rage clicks jumping from 9/day to 110/day across 23 users. The scout identifies the 'Pay now' button as the friction point, links example recordings, and files a P2 report routed to the checkout owner.
Quick Start
Ask the agent to run a session replay scout pass to check whether recordings are still being captured and whether any pages show concentrated rage-click friction.
