OPFS + Pyodide test harness
Simon Willison built a browser-based test harness combining OPFS (Origin Private File System) and Pyodide to explore whether Datasette Lite, a Python application running entirely in WebAssembly, can edit persistent SQLite files stored locally. The experiment, scaffolded by Claude Code for web, probes the intersection of client-side Python execution and browser file system APIs. This matters for the AI infrastructure layer: it tests whether Python-based data tools can operate offline and statefully in browsers without server backends, expanding where ML workflows and data applications can run.
Modelwire context
ExplainerThe less-obvious piece here is that OPFS grants browsers a sandboxed, persistent file system that bypasses the usual read-only constraints of WebAssembly environments, and Willison is specifically probing whether Pyodide can write back to SQLite files stored there across sessions, not just read them.
This is largely disconnected from recent activity in our archive. It belongs to a quieter but consequential thread in the broader client-side AI tooling space: the effort to make data-heavy Python applications run entirely in the browser, with no round-trips to a server. That matters for ML workflows because it shifts where inference, data wrangling, and even lightweight model evaluation can happen, moving them closer to the user's machine and away from cloud dependencies.
Watch whether Willison publishes a follow-up confirming that Datasette Lite can complete a full read-write-persist cycle on an OPFS-stored SQLite file without data corruption across browser sessions. If that works reliably in at least two major browsers, the serverless Python data tool case becomes substantially more credible.
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsSimon Willison · Datasette Lite · Pyodide · Claude Code for web · OPFS
Modelwire Editorial
This synthesis and analysis was prepared by the Modelwire editorial team. We use advanced language models to read, ground, and connect the day’s most significant AI developments, providing original strategic context that helps practitioners and leaders stay ahead of the frontier.
Modelwire summarizes, we don’t republish. The full content lives on simonwillison.net. If you’re a publisher and want a different summarization policy for your work, see our takedown page.