What problem does it solve?
Teams often run cron jobs that send $set identify events from their data warehouse into PostHog to keep person and group properties current. This Skill replaces that pipeline by mapping columns from a synced warehouse table or materialized view directly onto PostHog properties, on the table's existing sync schedule.
Core Features & Use Cases
- Whole-table column mapping: Proposes a complete
column_property_map in a single call, excluding identifier columns, identity properties like email, and sensitive columns, while flagging collisions with existing properties.
- Definition and source creation: Creates the custom property definition (person or group target) and binds it to a synced table (
external_data_schema) or materialized view (saved_query) with the chosen key column.
- Backfill and run verification: Triggers backfills for historical rows and reads the run funnel (
rows_read, changed, existing, produced) to diagnose mappings that appear to do nothing.
- Use Case: A user says "sync my postgres.customer_attributes columns to person properties" — the Skill finds the synced table, confirms the distinct ID column, proposes the mapping, creates it, and verifies the first run.
Quick Start
Map the columns of my synced Postgres customer_attributes table to PostHog person properties and run the first backfill.
