Editorial methodology

Editorial stance

The tone is measured, factual, and never alarmist. WHO has assessed the risk to the general population as “low” and that is the level we hold across the entire site, including in headlines. Emotionally charged words (“explosion”, “catastrophe”, “killer”) are avoided unless quoted directly with quotation marks and a link to the original source.

HantaTracker is not a health authority site. No individualised medical advice is given here. In case of doubt or symptoms, readers should contact their physician or emergency services.

Source hierarchy

When two sources disagree, the source higher in this hierarchy prevails. Every published figure must be traceable to a level 1 to 3 source.

1. Level 1 — Health authorities. WHO, ECDC, CDC, official national health authorities (Santé publique France, French DGS, Robert Koch Institute, NHS, BAG, etc.).
2. Level 2 — Peer-reviewed journals. NEJM, Lancet, JAMA, Nature, Science, BMJ, PLOS, Eurosurveillance.
3. Level 3 — Established press. Le Monde, BBC, Reuters, AFP, NYT, Le Parisien, France Info, NPR, Le Figaro.
4. Level 4 — Regional press. France 3 regions, Ouest-France, Midi Libre, BFM regional, identified local outlets abroad.
5. Level 5 — Summaries. Wikipedia, only as a summary referenced to level 1 to 3 sources already identified independently.

Publication workflow for a figure

Before a figure appears on this site, it goes through the following steps:

1. Identify the primary source through direct search on health authority websites or established press.
2. Verify that the URL actually contains the figure cited — no loose paraphrasing, no inference from a headline alone.
3. Date the source explicitly: day of publication or update.
4. Record in the site's structured data with the short source name and full URL.
5. If the figure cannot be found, it is not published. No figure is “estimated” or “inferred” without saying so, and even then, we prefer not to publish.

Explicit prohibitions

• Inventing a “plausible” figure based on general knowledge.
• Inventing a report name, publication identifier, or URL.
• Citing a figure via “according to a source” without naming the source.
• Adding a statistic without being able to produce the exact URL that contains it.
• Modifying an already-published figure without leaving a correction trail.

Automated updates

Every hour, an agent pipeline (public code) queries about twenty RSS feeds and Google News queries, deduplicates, verifies that each URL still resolves, and proposes new articles for integration. No main figure (cases, contacts, deaths) is ever updated automatically: these always go through human review before publication.

Hourly snapshots of the site's state (cases, contacts, affected countries) are stored in /cases-timeline.json for after-the-fact auditing.

Privacy and personal rights

A person is named on this site only if:

• they have spoken publicly (interview, statement, social-media post), or
• their name has been cited by an established press source or an official authority.

In both cases, the original source is linked. No individual health information is published without explicit declaration from the person concerned or an official source. Any takedown request sent to contact@hantatracker.fr is processed within 48 hours.

Copyright

The photos used on this site all come from Wikimedia Commons under a free licence (CC BY, CC BY-SA, public domain). Full attribution appears in the caption and in the central images.json registry. No press content is screenshotted: articles are cited by title and URL only. Social-media content is embedded through the platform's official embed, never as a screenshot.

Original HantaTracker texts are published under CC BY 4.0. You may reuse them freely as long as you credit: “Source: HantaTracker.fr (citing <primary source>)”.

Correction policy

Every correction of a published fact is publicly tracked, with date, field corrected, reason and new source. To report an inaccuracy: contact@hantatracker.fr. Response time: 48 hours.

Accessibility

The site aims for WCAG AA conformance. This covers: strict heading hierarchy, ARIA labels on all interactive elements, mandatory text alternatives for all images, verified WCAG AA contrasts, prefers-reduced-motion honoured, skip link and language declaration present.

Resources for journalists and researchers

• /outbreak.json — raw outbreak data (cases, contacts, deaths, clusters, hospital geolocations).
• /cases-timeline.json — hourly history since 11 May 2026.
• /graph.json — knowledge graph (entities and relations).
• /llms.txt and /llms-full.txt — summaries for AI assistants.
• /.well-known/agents.json — agent contact manifest and citation policy.
• github.com/rdelfosse/hantatracker-agents — public source code for the monitoring routines.

Author

The site is published in a personal capacity by Romain Delfosse. romaindelfosse.fr — LinkedIn — GitHub.

For editorial questions, reports, or takedown requests: contact@hantatracker.fr.