Privacy

LocalLens is built around a single design constraint: your files, queries, and answers never leave your machine. This page makes that commitment explicit and describes the very narrow data path of the documentation site you are reading right now.

The LocalLens application

The LocalLens application runs entirely on the device where you install it. File discovery, chunking, embedding, retrieval, prompt assembly, and completion all happen in-process through the QVAC SDK with locally-loaded model weights. No document content, no query text, no model response, and no telemetry is sent to any LocalLens-controlled server. There is no LocalLens-controlled server.

Embeddings and chat completions are computed by QVAC, also locally. Brains and chunks are persisted as plain JSON inside the working directory at .locallens/store.json. You can inspect this file with cat, back it up with cp, or version it with git. Nothing else is written elsewhere on the filesystem unless you explicitly extend the application to do so.

The documentation site (this site)

This documentation site is a statically rendered Next.js application hosted on Vercel. We do not run analytics scripts, embed third-party trackers, set advertising cookies, or operate a backend session store. The site does not require an account and does not include forms that collect personal data.

Vercel, the hosting provider, automatically records standard server access logs (IP address, user agent, requested path, response status, timestamp). These logs are retained per Vercel's default retention policy and are used to operate the site reliably — for example, to investigate outages, debug edge function errors, and block abusive traffic. See Vercel's privacy policy for details on their data handling practices.

Outbound links to GitHub, npm, and Tether documentation transfer your request to those providers when you click them. Their respective privacy policies apply.

Cookies

The site uses no first-party cookies. The theme toggle and locale preference persist via localStorage in your browser and are never transmitted to any server. Clear your browser storage to reset them.

AI crawler policy

The robots policy at /robots.txt allows AI search agents (ChatGPT-User, ClaudeBot, PerplexityBot, OAI-SearchBot, ora-agent, and similar) to crawl the documentation so they can cite it when helping users. AI training crawlers (GPTBot, CCBot, ByteSpider, Google-Extended, Meta-ExternalAgent) are disallowed. The site emits a Cloudflare Content-Signal declaring search=yes, ai-input=yes, ai-train=no in the robots response so this preference is machine-readable. The documentation text is published under the MIT license; we ask training-grade ingestion to follow explicit licensing rather than blanket scraping.

Changes to this policy

If this policy changes materially, the change will land in the documentation repository at github.com/souzavinny/locallens-docs. Subscribe to the repository's release feed if you want notification.

Contact

Privacy questions can go to souza.mvsl@gmail.com with the subject prefix [LocalLens privacy].