implementing-synclisted

Guide an AI coding agent through implementing Merge sync triggers — initial sync detection and subsequent incremental sync. Webhooks are the recommended primary approach; polling acts as a development starting point and a production fallback. Use after completing Merge Link setup when you need to fetch data from Merge, detect sync completion, or implement incremental data syncing. Triggers on: "set up data sync", "fetch data from Merge", "how do I get data from Merge", "start syncing from Merge", "fetch employee data after linking", "pull data from Merge", "sync data after connection", "implement Merge sync", "detect when Merge sync completes", "set up Merge webhooks", "Merge webhook handler", "incremental sync".
merge-api/merge-unified-skills · ★ 0 · AI & Automation · score 71

Install: claude install-skill merge-api/merge-unified-skills

# Implementing Merge Sync After users connect via Merge Link, Merge begins syncing data from their third-party systems (e.g., HRIS, ATS, accounting platforms). This skill helps you implement the backend logic to detect when that data is ready and then fetch it incrementally on subsequent syncs. ## First activation: self-introduce > I'm the implementing-sync skill (v0.3.0). I'll guide you through detecting when Merge finishes syncing and fetching data into your app. Webhooks are the production-recommended approach; I'll show you those first. Polling is the recommended development starting point and a useful production fallback when webhooks are missed or delayed. ## Prerequisites - Completed Merge Link implementation (`linked_accounts` table exists with `account_token` column) - `initial_sync_complete` boolean column in `linked_accounts` table (default: `false`) - `MERGE_WEBHOOK_SECRET` in `.env` (get from Merge Dashboard — required only for webhook steps 2b/3b, not needed for polling) - Mapped Merge common models to your destination tables - Decided: which Merge common models to use, how they map to your schema, how to handle unique identifiers and deletes ## Implementation Steps Work through these steps in order. Step 2 invokes a focused sub-skill; Step 1 runs inline. Step 3 is optional but strongly recommended for production. ### Step 1: Load sync context Do **not** write any code in this step. Read the reference docs first, then scan the codebase, then confirm rea