Maintained by Pablo Ruiz Quintero. Last reviewed . Every claim below is meant to be checkable against the site itself; where one stops being true it is removed from here first.
This page exists so that the claims elsewhere on the site can be checked against something. It is short on purpose: every line below is either true today or should be deleted.
Who is behind this
One person: Pablo Ruiz Quintero, who builds the data pipeline, sets every editorial guideline, the system that verifies the data and the minimum standards that decide whether a page is published, reviews pages for structure, approves what goes live and keeps the verification log. The responsibility for what is published is his. He does not write every sentence on the site: most of the written text is drafted with AI, as the next sections explain. There is no anonymous content team, no outsourced articles and no invented author. Pages carry his name in their structured data as the person responsible, pointing at a single identity with a public ORCID record.
He is not a fisheries biologist, and the site does not claim otherwise. The expertise being offered here is in reading public instrument data correctly and refusing to publish what cannot be verified. Biological information comes from state fisheries agencies and published literature, and where it has been wrong on this site, the log says so.
Where the numbers come from
Conditions come from four instruments, each named with the endpoint queried on the methodology page: USGS Water Services for flow and stage, NOAA CO-OPS for tides, Open-Meteo for weather and pressure, and an astronomical calculation for solunar windows. Licence fees come from the agency that sells the licence, with the date of reading recorded on each state page. Where a source is a publisher acting for the agency rather than the agency itself, the page says which.
We do not use compiled fee tables or search-result summaries as a source. Three states are listed wrongly in those, which is recorded in the log.
How the text is made: AI, checks and a person
Being straight about this matters more than it is comfortable. The daily report on each water page is generated by software from that morningโs readings: the activity score, the verdict line and the timing windows are computed, not typed. That is the point of them, and it is why they can be regenerated every day for 996 waters without anyone inventing a fishing report.
The written text โ what each water is like, how a species behaves, what a licence covers โ is drafted with AI (Anthropicโs Claude) from the official pages of the agencies, water authorities and park services that manage each water. It is then checked against those pages by automated verification: separate AI checks and scripts that compare each fact with the cited page, compute dates from the published rules for two years, and test every page before it ships. On the newer pages (Australia, Canada, Japan, the season and licence pages) anything that could not be verified was taken out, and each page lists the official pages it was checked against and the date. The 996 US water guides were written earlier and do not yet list their sources; that audit is under way.
The personโs part is to set every guideline, impose the data-verification system and the minimum standards for what is and is not published, review pages for structure, and approve and decide what goes live; he answers for all of it. The AI is the executor: it drafts and checks within those rules, and it does not decide what is published. He does not proofread every sentence of every page, and we say so because a byline that implied otherwise would be the kind of claim this page exists to prevent. We work this way because it is the only way one person can keep about two thousand pages current with live data and named sources.
Each page says how it was made, next to its sources. The assistant on the site answers from the siteโs own pages and is constrained to link only to pages that exist, after it was caught inventing a URL. When a check or a reader finds an error, it goes in the verification log with the date.
How this site pays for itself
There is no paywall and no subscription. Since 29 September 2026 the site carries advertising from Google AdSense: the ads are chosen and served by Google, are marked as ads, and no advertiser can buy a place in the text, a ranking or a verdict. There are no affiliate links on licence pages: the buy link always goes to the state agency, never to a reseller, and if a site charges a processing fee on top of the state fee we say so. Should that change, this section changes first and the change is dated in the log.
What we will not publish
- No invented reports. Every daily report is generated from live readings taken that morning. There is no stock text waiting for a day when the API fails: if the data does not arrive, the page says so and refreshes.
- No fake reviews or catches. Angler reports on this site come from anglers. We do not seed them, write them or pay for them, and there are no invented star ratings anywhere on the site.
- A gauge is never assumed. A water only gets a live gauge or tide station after the API has been queried and the name it returns has been read. Six candidate gauges have been rejected for returning a different river.
- A figure with no source is not published. Fishing licence fees are read off the agency that sells the licence, with the date recorded. Three states are missing from the site entirely because their fee schedule could not be read.
- Mistakes are logged, not quietly fixed. The verification log records what was wrong, how wrong it was, and when it was corrected, including the errors that were ours rather than the sourceโs.
- No fishing or safety advice beyond the data. These are conditions reports and planning tools. Ice thickness, water safety and what is legal to keep are decisions that belong with you and the agency that regulates the water.
Corrections
Corrections are logged, not quietly applied. When a figure changes because it was wrong, an entry goes in the verification log with the date, what was wrong, what it would have cost the reader, and what was done. The log currently holds 65 entries, 45 of them our own errors.
To report something: hola@prqdigital.com. Include the page and what you believe is wrong. If you are an agency and we have misread your publication, that is the fastest way to have it fixed.