Where your data goes
Choose the OpenThunder surface that fits the work you want to do. Local analysis, the browser sample, Cloud, and AI actions have different data paths.
Browser diff sample
The diff sample processes pasted text in the page. It does not call a model, upload the diff, run tests, or inspect the rest of your repository. The page has no analytics script and blocks network connections from its JavaScript.
The tool does not save a draft. Clear input & results removes them from the page. Your browser and clipboard are separate: clearing the tool does not erase text you copied. Reports include file paths, so read them before sharing.
Try the built-in synthetic example if you do not want to use your own code.
Desktop and CLI
Local repository analysis runs on your computer through the local engine. You can begin without an account or model key. Generated reports and caches can contain file paths and other repository information; treat them as part of your project data.
Local analysis does not mean every feature is offline. Sign-in, updates, integrations, Cloud publishing, and AI actions can make network requests. A configured remote AI provider processes the prompts and selected context sent for that action.
OpenThunder Cloud
Cloud is a hosted workspace and requires sign-in. Cloning a repository in Cloud fetches it onto the hosted service. Uploading a folder sends selected text files and their paths to that service. It is different from publishing a locally generated report.
Choosing a folder first reads it into the browser for review. No source is sent by that selection alone. Review the file list, uncheck files, and explicitly confirm before uploading. The preview is held in page memory, not saved as a browser draft. Discarding it does not delete the original folder or revoke a completed upload.
The folder picker filters common build output, binary extensions, and files over 1 MB. Reading is capped at 4,000 files or 25 MB and may be partial. It is not a secret detector or a guarantee that credentials have been removed. Use a prepared copy with secrets removed, or choose Desktop when the source must remain on your machine.
Cloud scan sandboxes have a scheduled cleanup mechanism. Do not treat that as a guarantee that every copy, log, or derived result disappears on a fixed timer. Retention and deletion depend on the feature and the records involved. See the Privacy Policy, or contact support@skillstechtalk.com for an account-data or deletion request.
AI providers and training
The browser diff sample does not use AI. For model-backed features, the provider and configuration determine which prompts and context leave the device. An OpenThunder account or a local installation does not, by itself, establish a provider's retention or training settings. Read the provider's terms and your account settings before sending private material.
Our Privacy Policy describes Skills Tech's data-use commitments. This FAQ does not make a blanket no-training promise on behalf of every third-party provider.
Public-site analytics
Optional Google Analytics on this public site stays off until you allow it. Keep analytics off leaves the tools and downloads available. A Global Privacy Control or Do Not Track preference keeps it off. Use Analytics choices to change your selection later.
Our explicit events count broad page categories and entry clicks, such as Desktop, Cloud, or the browser sample. They do not include pasted code, answers, repository names, raw destination URLs, incoming query parameters, or URL fragments. A download click is not an installation or a completed review. The browser diff sample has no analytics.
Your analytics and appearance choices are saved in this browser. Cloud sign-in uses separate account storage and cookies.
Cloud app analytics
The OpenThunder Cloud web app (app.openthunder.ai) asks first, the same way this site does. Google Analytics does not load, sets no cookies, and sends nothing until you choose Allow analytics. Keep analytics off is just as easy and every feature works with it. A Global Privacy Control or Do Not Track preference keeps it off without asking. Use Privacy choices in Settings or on the sign-in screen to change your selection; turning it off stops sending at once and removes the analytics cookies.
When allowed, it counts which pages of the app are visited by page type, such as Settings or Repositories. It never receives your email, account or user identifier, repository names, file paths, code, or the full web address. IP addresses are anonymized, and advertising features and Google Signals are off.
Desktop app usage sharing
The desktop app sends no usage data unless you turn on Share which features you use in Settings. It is off by default. When on, the app sends only the name of each feature you use (for example, opened a repository or ran a Check), the app version, and the day. It never sends code, file names, repository names, prompts, or an identifier for you or your computer. Turn it off in Settings at any time and sending stops at once. Local Only mode turns it off whatever you chose. The desktop app has no Google Analytics.
The command-line tools send nothing unless you run openthunder telemetry on (or set OPENTHUNDER_TELEMETRY=1). openthunder telemetry shows exactly what is sent.
Getting started without private data
- Try the browser sample.
- Review the recorded synthetic example.
- Download Desktop for local analysis.
- Read about Cloud before connecting a repository.