Cloud//Cloudflare//Wrangler//.wrangler

Hidden directory that Wrangler creates in the project root. Contains the local state of all Cloudflare services — the "local cloud" on your machine.

Hidden directory that Wrangler creates in the project root. Contains the local state of all Cloudflare services — the "local cloud" on your machine.

The D1 SQLite files live here. When you write to D1 locally (--local), the rows go into .wrangler/state/v3/d1/.../*.sqlite. That IS the local database — no separate server, no daemon, just a file.

wrangler d1 export --local reads from these SQLite files and generates a dump.sql — the transport file you can then execute against the remote DB. The data flows: .wrangler/ → dump.sql → remote D1.

Workers KV local data, R2 local buckets — all simulated here too. Same pattern: local files that mimic the cloud service.

Lives in .gitignore. GitHub never sees it. Code deploys (GitHub Actions, wrangler pages deploy) push dist/ + functions/ only — .wrangler/ stays on your machine.

If you delete .wrangler/, you lose all local data. Next wrangler dev recreates the directory structure but the data is gone. Pull from remote to restore.

No config needed — Wrangler creates and manages it automatically based on wrangler.toml bindings.