Powersync

This teaches Claude how to set up PowerSync, an offline-first sync framework that replicates data between SQLite and backends like Supabase, Postgres, MongoDB, or MySQL. It's built around the PowerSync CLI and enforces a disciplined workflow: ask about Cloud vs self-hosted first, configure the backend before touching client code, always confirm the target instance before deploying. The skill defaults to read-only help on existing projects unless you explicitly authorize infrastructure changes, which is smart guardrailing for production systems. Covers sync stream configuration, client SDK setup across eight platforms, upload queue mechanics, and debugging. Best used when you're wiring up local-first architecture and want to avoid the usual config file mistakes and deployment footguns.

skills.sh installs: 1.4K
GitHub stars: 22
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add powersync-ja/agent-skills --skill powersync --agent claude-code

Installs into .claude/skills of the current project.

Files

PowerSync Skills

Use this skill to onboard a project onto PowerSync and to keep it healthy afterwards, without trial-and-error. Treat this as a guided workflow first and a reference library second.

Agents: Read AGENTS.md before proceeding. It contains the mandatory compliance rules and onboarding playbook. The Quick Rules below are a reminder, not a substitute. powersync login is PowerSync Cloud only (PAT); self-hosted does not use it.

Terminology

Operator — the human directing this agent (whose request you are fulfilling).
User — an end-user of the operator's PowerSync app (JWT subjects, the row a sync stream filters by, the person calling disconnectAndClear()).

If a sentence is ambiguous, default to the operator interpretation. Full legend in AGENTS.md.