report-automation
Installation
SKILL.md
report-automation
Core Philosophy
Manual executive reporting—exporting CSVs from 6 different SaaS dashboards, pasting them into Excel, manually aligning pivot tables, and taking screenshot crops for Slack—is a colossal waste of engineering and operational talent. Effective report automation builds a reliable, idempotent data pipeline: automated API extraction, schema normalization, anomaly detection, deterministic markdown/visual generation, and scheduled multi-channel distribution.
4-Step Report Automation Engineering Pipeline
Step 1: Multi-Source Data Ingestion & Extraction
- API Connectors vs Direct Database Replicas:
- Extract raw metrics directly from source systems via read-only SQL replicas or official REST/GraphQL APIs (Stripe, GitHub, Google Analytics, PostHog).
- Use deterministic pagination and incremental sync windows (e.g.
updated_at >= NOW() - INTERVAL '24 HOURS').
- Idempotency & Caching:
- Cache intermediate raw JSON/CSV responses in a local data lake or object store (S3/MinIO) before transformation to allow re-runs without hitting external API rate limits.