Wind briefing
Editorial standards
How the daily wind briefing is curated, what we filter out, and how to flag errors. The briefing publishes under the Turbit brand but reads as independent industry curation — these are the rules that make that claim true.
What this briefing is
A weekday digest of wind-energy news for operators, asset managers, and insurers. Each edition surfaces 2–5 items chosen for signal density: policy and regulation, market data, operational risk, insurance and loss trends, repowering decisions, fleet incidents reported by trade press. We under-weight stock and M&A noise, single-project construction milestones, and feel-good corporate announcements.
How items are selected
Every weekday at 06:00 UTC, our pipeline pulls the most recent 72 hours of items from a fixed allowlist of trade-press RSS feeds and Google News queries (per locale). A language model summarises each item under a strict schema: every published item must trace to one URL in that allowlist; URLs that don't appear in the input feed are dropped automatically before publish. The model also emits a category and a source-tilt label which the reader sees alongside each item.
Source allowlist (current edition):
- Prensa especializadaEnergías Renovables
- Prensa especializadaEl Periódico de la Energía
- Prensa especializadaREVE
- Voz de la industriaAEE — Asociación Empresarial Eólica
- AgregadorGoogle News (ES)
Editorial rules
These are encoded in the prompt and enforced by the publish schema:
- No advocacy framing. Source claims are reported, not endorsed ("X argues Y" rather than "Y").
- Every item traces to one input URL. Hallucinated URLs are filtered before publish.
- No Turbit mentions. The briefing does not name Turbit, its products, its competitors, or any predictive-maintenance vendor unless that vendor is named in a source headline.
- No OEM PR amplification. Vendor product launches without independent reporting are dropped.
- Source tilt is shown. Each item carries the source's tilt label (trade press, industry voice, regulator, insurer data, aggregator) so readers know the perspective.
AI disclosure
Headlines, summaries, and the editorial intro are written by Anthropic's Claude (Sonnet) under the schema and rules above. No human reviews each edition before it publishes; the schema-level filters (URL traceability, category coercion, advocacy-keyword sanitisation) are the only guards. Every page declares this in `schema.org` JSON-LD via `creativeWorkStatus: "AI-generated"` and lists the model as the `creator`. Always click through to the cited source — the briefing is a starting point, not a primary source.
Corrections
Spotted a misattribution, a hallucinated URL, or a source we should drop? Email briefing@turbit.de with a link to the edition and a one-line description. Corrections are applied within one business day; structural changes (source allowlist, prompt rules) are recorded here.
Retention and republishing
Every edition is archived permanently at a stable URL. The cron may re-run on the same calendar day if a fetch failed; that overwrites the day's edition with the corrected version. Older editions are not re-edited after the day they publish — corrections appear in newer editions or here on this page.
Managing editor
The briefing's source allowlist, neutrality rules, and editorial schema are designed and maintained by Michael Tegtmeier (co-founder, Turbit Systems GmbH). Day-to-day editions are not personally reviewed; the role is to own the methodology and the corrections process.
Updated every weekday at 06:00 UTC.