Features

Analytics

Connect a PostHog project and query events, persons, and sessions with HogQL — no separate analytics tool required.

Analytics

Dora connects to a PostHog project over the read-only HogQL Query API — there's no wire protocol to implement, so the connector works the same way against PostHog Cloud (US or EU) as it does against a self-hosted instance.

Connecting a project

Add a PostHog connection the same way you'd add any other data source, using a personal API key with query:read scope. See Connect → PostHog for the credential setup.

Browsing the schema

PostHog has no information_schema, so Dora reflects each virtual table's columns by running a zero-row SELECT * and reading back the HogQL Query API's columns/types metadata. The tables surfaced in the schema browser are:

  • events
  • persons
  • sessions
  • groups
  • cohort_people
  • raw_sessions

Row counts shown next to each table are a live SELECT count() — they're hidden rather than shown as zero if the count can't be read.

The built-in dashboard

Opening a PostHog connection adds an Analytics entry to the sidebar. It runs six HogQL queries against the connection and renders them as a dashboard:

PanelWhat it shows
KPI tilesTotal events, unique users, and pageviews over the last 7 days
Monitored sitesHosts sending events in the last 30 days, with per-site event/user/pageview counts
ActivityDaily events and unique users over the last 14 days
Top eventsThe 12 most frequent event names in the last 7 days
Top pagesThe 10 most-viewed URLs ($pageview events) in the last 7 days
Top browsersThe 8 most common $browser values in the last 7 days

Every query accepts an exclude localhost filter that drops traffic from localhost, 127.0.0.1, 0.0.0.0, and [::1] hosts, and a manual refresh re-runs all six.

The dashboard is read-only, matching the connector's write-blocking behavior for PostHog: Dora never mutates PostHog data, only queries it.