Reviewed 2026-05-29
Source Integrity & Editorial Policy
What we publish, how we verify it, and how we handle the times we get it wrong.
Plain-English summary
- We are not an official Sullivan County, City of Bristol, City of Kingsport, or State of Tennessee tourism site. We are independently built and locally maintained.
- Each site detail page shows a
Last verifieddate. That is the date a human checked the hours, prices, address, and contact information against the operator's own material. We are extending the same stamp to tours, events, and narratives. - We do not use AI to generate historical detail. We do not use AI-generated photography. Heritage tourism only works if visitors can trust what they read.
- We do not run advertising and we do not place tracking pixels. Lodging "Check rates" buttons use a UTM tag so the county can see the platform is working — we do not see your booking.
- What we do store, only when you opt in:
- Weekly digest signups — your email address, the source page you signed up from (e.g. "homepage-footer"), and the date. Stored encrypted at rest in Cloudflare KV. Used only to send the Friday digest. Unsubscribe at
/api/digest/subscribe?email=you@example.comwith the DELETE method, or reply to any digest. - Push notification subscriptions — the opaque endpoint URL your browser/OS generates plus the two public-key fragments the Web Push protocol needs. Stored in Cloudflare KV, keyed off a hash of the endpoint. Used only to deliver Sullivan County alerts you opted into. Revoking notification permission in your browser disables it; you can also DELETE the subscription via the API.
- Anonymous Companion queries — when you use Ask the Companion we log the query text + the IDs of the chunks we returned (so we can improve relevance). We do NOT log your IP, browser, or any cookie. Logged in Cloudflare Workers Logs with a short retention.
- API request logs — every
/api/*call writes one structured log line: method, path (no query string), HTTP status, response time, and a short request id. No IP, no cookie, no PII. Used for cost + reliability monitoring. - No third-party trackers: no Google Analytics, no Facebook pixel, no advertising SDKs.
- Weekly digest signups — your email address, the source page you signed up from (e.g. "homepage-footer"), and the date. Stored encrypted at rest in Cloudflare KV. Used only to send the Friday digest. Unsubscribe at
- If you spot an error, email corrections@wheretennesseebegan.com. We aim to acknowledge within one business day and publish a correction within 72 hours when the claim is wrong.
How we verify (today)
For every site we publish, we record hours, admission, contact info, accessibility notes, and parking from the operator's own published material — their website, brochure, or signage — at the date recorded in the entry's lastVerified field. All 59 sites currently published carry a verified date.
What we are still building: a 90-day rolling review cycle, a public freshness report, and the same lastVerified stamp applied to tours, events, and narratives. The helpers and report exist (see scripts/verification-report.ts); the schedule and the broader stamping have not landed yet. We do not auto-bump verification dates without a human review.
How we cite (the commitment, and where we are)
Heritage narratives — the long-form stories under /stories — are committed to citing primary sources where available (state archives, county records, museum catalogs, peer-reviewed scholarship) and reputable secondary sources otherwise (Tennessee Encyclopedia, academic histories, Smithsonian publications).
What is true today: a sources field exists on the narrative schema for citation entries, and a reviewedBy field for the named external reviewer. What is still in progress: retroactively adding sources to the eight narratives currently published, and rendering them as visible footnotes on the page. We will not call a narrative "cited" until it actually is.
Wikipedia is used as supplementary context for site descriptions. The Wikipedia article title is stored in wikipediaTitle and the enrichment data is loaded from wikipedia-enrichment.json. Any visible Wikipedia-sourced text is attributed to its source article in line with CC BY-SA.
Photography & licensing
Every photograph that appears on this site is either licensed for use, in the public domain, or commissioned with documented permission. We never use AI-generated photography and never present a generic Appalachian stock image as a specific Sullivan County place.
The full third-party license file is at LICENSES.md in the repository.
Inclusive heritage
Sullivan County's heritage includes Cherokee homeland from before 1776, Black settlers and tradespeople, the women of the Watauga Association, and frontier families from many backgrounds. We commit to representing all of them, not just the most familiar accounts. Periodic audits of the breadth of voices in our content are tracked at docs/reports/inclusive-heritage-audit-*.md.
What we will never do
- AI-generated historical facts in narratives.
- AI-generated photography passed off as real.
- Unverified claims about hours, prices, or accessibility.
- Quote a person without their permission.
- Sell, share, or monetize visitor data.
- Take a payment in exchange for editorial placement.
Corrections
Email corrections@wheretennesseebegan.com with the page URL and what you'd like changed. We acknowledge within one business day and publish corrections within 72 hours for verifiable factual errors. We update the affected page's lastVerified stamp at the same time so the change is auditable.