This is your Nim codebase — like a VS Code working folder, stored in your browser (IndexedDB). Click a file to edit it. The whole folder is mounted at /workspace on every compile, so import sibling and import sub/mod work. Run compiles the file open in the editor; Build compiles every .nim file into its own .wasm in the Site folder; the site boots the build entry (main.nim, or the example you picked — use a file's entry button to choose another). Export ZIP downloads the whole folder.

This is the deployed static website — the working folder of the app pane, which renders this folder's index.html. Press Build to compile your project into it (one .wasm per .nim file + index.html with the JS runtime inlined, minified; the entry's wasm is the one index.html boots). Export ZIP downloads the whole folder. You can add your own files (CSS, images, …) — Build only writes generated .wasm files and a generated index.html (a hand-edited index.html is kept, and only then does Build write nim-runtime/* next to it — legacy mode), and it removes stale .wasm files that no longer match a project .nim source — an uploaded .wasm matching no project module is removed on Build; non-wasm uploads are never touched. Built .wasm files are obfuscated (Nim name/producers sections stripped, decoy GCC comment injected), and calling bindDomain("example.com") in your Nim code makes the deployed site refuse to boot on any other domain — deterrence against casual theft, not cryptographic security.