This document describes how Level Up Adventuressources, drafts, fact-checks, and corrects its coverage. It is intentionally specific. If a future post or video appears to violate one of these rules, the rule is the standard — the post is the bug. Email support@level-up-adventures.com with the URL.
§ 01What we publish
- Weekly digests. Theme Park Pulse (crowds, closures, events, new rides) and Family Deal Drops (all-inclusive, cruise, theme-park ticket, flight deals). Built from public APIs and merchant promotion pages.
- Destination field guides. First-person reporting from trips the operator has taken with family. Time-stamped photo and video evidence is retained for any claim that a place is a certain way at a certain time of year.
- Travel-day mechanics.Logistics: transit, eSIMs, weather windows, where you need a car, where you don’t. Sourced from operators’ own pages and first-person experience, not from press releases.
§ 02Sources we use
- Public APIs. queue-times.com for live theme-park wait times. National Weather Service and equivalents for weather. Public CDC and State Department travel notices for safety information.
- Public web pages.Park news pages, merchant promotion pages, and aggregator special-offers pages, extracted via Crawl4AI markdown parsing with a per-host throttle (typically 1.5–2 seconds between requests) and 429 backoff handling. We honor
robots.txton every request — if a source disallows our user-agent, we don’t scrape it, and we say so in the post’s “source issues” section. - First-person reporting.The operator’s own trips, with photo / video timestamps, transit receipts, and reservation records preserved off-site. Field guides cite which observations are first-person and which are sourced.
- What we do NOT use. Press-release-driven content, pay-for-coverage syndicated feeds, or AI-generated descriptions of places no human has actually been. If we cover a place, we either went there or we sourced the claim to an identifiable public-page citation that the reader can click.
§ 03How AI is used (and not)
— Plainly stated, because most disclosures aren’t.What AI does. AI tools draft and edit prose from already-sourced material we scraped or wrote. They reorganize bullet points, smooth headlines, and produce the first pass of explanatory copy that a human operator then edits, fact-checks against the source links, and signs off before publication.
What AI does not do.AI does not invent claims, prices, ride status, or destination details that aren’t already in the source material we collected. AI output is treated as an untrusted draftthat has to be verifiable against the cited sources. If a claim in a post can’t be traced back to a source link, it doesn’t ship. Hallucinated content is a bug we treat as a correction obligation, not a stylistic choice.
Where AI is disclosed in posts.Each weekly digest ends with a “How this digest was built” section that names the extraction tooling, the sources, the throttle settings, and any source errors that were skipped. Field guides — AI-drafted or hand-written — are governed by this page as the standing disclosure. We don’t append per-post “built with AI” lines; they add review-ack noise without changing the editorial discipline that’s already stated above (sources still cited inline, hallucinations still corrected, untrusted- draft posture still applies). Posts whose drafting process differs materially from what’s described in this section — e.g., a guest contributor with a different review pipeline — carry a per-post note in their byline.
Deal scores and buy/skip guidance.The weekly Family Deal Drops attaches editorial scores (family fit, value, urgency, confidence) and a plain “book now / consider / compare first / skip” recommendation to each deal. Those scores and recommendations come from transparent rules plus AI-assisted summaries, and a human operator reviews and approves every one before it publishes. We never assert a savings figure the source doesn’t support — when public detail is thin, the post says so rather than guessing.
Hero imagery may be AI-generated.Some posts use an AI-generated illustration as their lead hero image. We reach for AI hero art when we don’t have a first-person photograph that does the post’s subject justice — an interior architectural moment we witnessed but couldn’t cleanly capture, a scene that includes a child whose face is governed by § 08, or a visual subject that doesn’t yield well to phone photography. AI hero imagery is illustrative, not documentary: we do not use it to fabricate a scene we didn’t witness, to misrepresent crowds, weather, conditions, or appearance of a destination, or to imply we were somewhere we weren’t. The same untrusted-draft posture from the paragraphs above applies — the operator reviews each rendered image against first-person memory before it ships. Like AI-drafted prose, AI hero imagery is governed by this standing disclosure rather than per-post credit lines, so the post page itself reads cleanly. Public-domain or licensed stock imagery, when used, is credited inline on the post.
§ 04Fact-checking and pre-publish review
Before any digest publishes, the operator verifies a sample of cited links resolve, a sample of price / status claims match the linked source, and that no figure has been altered between source and post. Field-guide posts are reviewed against retained first-person evidence (timestamped photos, receipts, transit logs).
Travel content can go stale fast. We mark every digest with a publication dateand we don’t back-date or silently update prices. If a price changes after publication, the post still shows what was true at publication, and we add a corrections note — not a stealth edit.
§ 05Preview discipline
Previews are not summaries. The excerptfield, the OG card, the Twitter card, the homepage teaser, the blog-index lead — none of them resolve the post. The reader’s job at preview-scan is to feel a pull. If the post’s payoff is a specific card you should buy, a specific neighborhood you should skip, or the one transit move that runs the trip, the preview does not name it.
Two principles:
- Set up the question; let the post be the answer.If the post answers “what’s the one transit move that runs a Barcelona family trip?”, the preview poses the question. Naming the Hola BCN card up front kills the click.
- Specificity, not curiosity bait.Generic teasers (“you won’t believe what we found”) are clickbait. Specific gaps (“the one transit move that saved $200 across three days”) are editorial craft. The gap should be something a reader plausibly wants to close, not artificial vagueness for its own sake.
This rule applies to every surface the post is summarized at: in-site excerpts, social cards, search snippets where we can influence them, email newsletter teasers, and YouTube descriptions when the description doubles as a written summary. It also rules out one machine-generated tic: AI-confidence boilerplate (“Sources verified,” “No issues to flag”) at the end of a post — it reads as review-ack and undermines the editorial voice. If there are genuine caveats, we put them in the corrections note; if there aren’t, we say nothing.
§ 06Affiliate independence
Editorial decisions — what gets covered, what gets ranked, what gets called good or bad — are independent of any affiliate relationship. Specifically:
- We do not accept payment to feature destinations, products, or merchants.
- We do not accept comp’d trips, comped tickets, or sponsored stays in exchange for coverage. If we ever do, the post will say so in its first paragraph, not its footer.
- We cover deals from non-affiliated sources in Family Deal Drops when those deals are better than the affiliated alternatives, and we mark them clearly as non-affiliate.
- We publish negative or skeptical coverageof merchants we have affiliate relationships with when the underlying product earns it. The relationship is with the network, not with the merchant’s reputation.
See the Affiliate Disclosure for the full list of affiliate networks and merchants. The disclosure document is the single source of truth for what we earn from.
§ 07Corrections policy
We correct material errors. “Material” means a factual claim a reader could plausibly act on (a price, a closure date, a safety statement, an attribution). Typos and copy-edits are fixed silently. Material corrections are noted at the bottom of the post with a timestamp and a one-line description of what changed and why.
To request a correction, email support@level-up-adventures.comwith the post URL and the specific claim that’s wrong. We respond, even when we don’t end up making the edit.
§ 08Off-limits topics
- Children’s privacy.We do not publish identifiable photos of children we did not raise without explicit, written, contemporaneous parental consent. The operator’s own children appear in our content with parental consent — sparingly, never as the focal subject of a piece, and never used to make a destination or merchant look better than it actually is. Where a face would be identifying in a context that doesn’t need it (background-prominent shots, hero images, social-thumbnail surfaces), we obscure or substitute it — and where a child’s face has been replaced with an AI-generated face for privacy, the post or video discloses that.
- Travel safety. When CDC, State Department, or equivalent national authorities have issued an active advisory for a destination, we link the advisory and we do not tell readers to disregard it.
- Medical, legal, financial advice.We are not a doctor, a lawyer, or a tax professional. Posts that touch travel insurance, visa rules, or medical preparation point to the relevant authority and don’t pretend to be definitive.
- Crisis exploitation.We do not publish “deals” coverage of a destination during an active disaster, evacuation, or human crisis affecting that destination.
§ 09YouTube and video coverage
Video coverage on the Level Up Adventures YouTube channel is held to the same sourcing, AI-use, fact-check, corrections, and affiliate-independence standards as written posts. If a video and a written field guide disagree, the version with first-person reporting wins. Video descriptions disclose any affiliate links in the same place an FTC reader would expect to find them.
§ 10Changes to this document
If we change these standards — particularly to add a new sourcing technique, a new AI tool, or a new disclosure requirement — we update the Last updated date and keep the previous revision available on request. Material changes are noted in the next weekly digest.