What problem does it solve?
PostHog's Inbox clusters raw product signals (errors, rage clicks, session replays, integration events) into actionable reports, but knowing which reports matter, whether they are trustworthy, and how to act on them requires a structured triage workflow. This Skill guides an AI agent through listing, filtering, drilling into, resolving, and implementing fixes for Inbox reports without assuming the inbox is even configured.
Core Features & Use Cases
- Inbox triage and prioritization: List reports filtered by status, priority, source product, or keyword, leading with reports where the current user is a suggested reviewer based on GitHub commit authorship.
- Report lifecycle management: Resolve, dismiss, or snooze reports individually or in bulk (1-100) with canonical dismissal reasons and notes that steer future scout behavior.
- Verified report implementation: Turn an actionable report into a PR by first validating the LLM-generated diagnosis against the actual codebase, then linking the PR back to the report.
- Source configuration review: Inspect which signal sources (error tracking, session replay, GitHub, Linear, Zendesk) feed the inbox and toggle them on or off.
- Use Case: A user asks "what should I look at first?" The agent checks source configs, lists ready reports, groups them by suggested-reviewer status and priority with deep links, then offers to drill into the top item and verify its diagnosis before opening a fix PR.
Quick Start
Ask the agent to show what is currently actionable in your PostHog inbox and summarize the top reports by priority.