Your timer stays in your browser
When you save a timer, NextReset stores the service, allowance type and reset timestamp in your browser’s local storage. The timer form does not send those values to our server. Use “Clear timer” to remove the saved entry, or clear this site’s browser data.
Hosting and page requests
Cloudflare hosts this website and handles page requests. Network information such as an IP address and browser request details may be processed to deliver and protect the site. See Cloudflare’s privacy policy.
Fonts and external links
This site loads fonts from Google Fonts and Tibo’s public avatar from X’s image service. Those services receive browser requests when their resources load. Links to public announcements and official help pages open services with their own privacy policies.
Forecast requests
The homepage requests a public forecast from our Cloudflare endpoint and refreshes it while the page is visible. The server reads a public reset-signal feed and OpenAI service-status data. Your account usage, personal timer and email are not sent to those feeds.
Public health and release monitoring
Our server periodically reads public OpenAI and Anthropic incident feeds and their official GitHub release metadata. The dashboard stores source-linked public records; it does not send your private conversations, project files or account usage to these sources. A healthy incident feed is not a measurement of your model quality. New incident and release notification categories are opt-in.
Codex quality measurements and visual preferences
We run our own fixed public tasks through a Codex subscription account. Public receipts contain task prompts, answers, grades, model configuration, timestamps and token counts. They do not contain our credentials or your conversations. The browser cookie described below also limits a pelican comparison to one preference per browser. We store the comparison identifier, preference and time for up to 90 days; public counts are anonymous. A local-storage entry remembers the latest preference. These votes express visual preference, not a scientific quality score.
Community picks
We use an essential, first-party HttpOnly cookie named nr_voter, lasting up to one year, to remember this browser’s pick and notification preferences. The server stores that random identifier, your chosen percentage or reset date, ballot and submission time. Votes are kept for up to 90 days; public summaries show counts and scores, never browser identifiers. Clearing cookies loses your ability to retrieve the old pick. One browser does not necessarily mean one person.
Browser notifications
If you opt in, we store your browser’s push endpoint, its encryption keys and your selected event types. Your browser’s push provider carries the encrypted notification. A delivery receipt lets us distinguish a provider accepting a message from this browser receiving it. “Turn alerts off” removes your subscription and cancels pending messages. Completed delivery records are retained for up to 30 days for troubleshooting. Previously accepted notifications may still arrive. We do not collect email addresses in this version.
Abuse prevention and aggregate counts
Write requests use a server-salted, daily IP hash for temporary rate limits; raw IP addresses are not stored in the application database. Rate-limit records expire and are pruned daily. Cloudflare may separately process network information under its own policy. We count actions such as saved picks, share-button use and notification receipts by day, separately from the aggregate web analytics described below. Public forecast snapshots are retained for 90 days.
Aggregate website analytics
We use Plausible Analytics to understand pageviews, referring sites, campaign tags, approximate country and device or browser categories. Plausible receives browser requests to provide these aggregate reports. We do not send your personal timer values, account usage, conversations, API keys or notification endpoint to Plausible. See Plausible’s data policy.
No account required
There is no account or payment system and no advertising script. RSS readers retrieve the public feed independently. Google Search Console is used to understand search discovery; it does not add a tracking script to these pages.
Calendar reminders
The calendar file is generated in your browser. If you import it into a calendar service, that service handles the reminder and its data under its own settings and policies.