What problem does it solve? Designing an identity source-of-truth to Okta sync pipeline (e.g. Entra ID) in Okta Workflows is full of hidden traps: silent type-sensitivity bugs, undocumented record caps per card, and async loops that flood API rate limits. This Skill consolidates the cross-cutting rules, key findings, and caveats so you avoid production failures before they happen. ## Core Features & Use Cases - Architectural guardrails: Enforces the core rule of never running bulk/looping work synchronously in a parent flow, using Stream Matching Records or For Each with bounded concurrency into helper flows. - Documented quirks and caps: Covers roughly 30 documented quirks including the 3,500-row Tables Search Rows cap, 4,000-group and 900-member Entra connector caps, and case-sensitivity pitfalls. - Platform change tracking: Summarizes 2025-2026 Okta Workflows changes such as the Okta ITP connector, Send Slackbot Message deprecation, and Connector Builder Polling Monitors. - Use Case: When starting an Entra ID to Okta sync design, ask for the field guide to get the TL;DR, five key findings, design-phase recommendations, and guardrail thresholds (280,000 events/day warning, 400,000/day cutoff) before writing any flows. ## Quick Start Ask the assistant to walk you through the Okta Workflows field guide rules for designing an Entra ID to Okta user sync pipeline.