Tools that look closely at things almost nobody looks at. Everything is built on open data — satellite imagery, orbital elements, public channels. No closed sources, and no guesses passed off as facts.
Map of Darkness
Nighttime lights from NASA Black Marble across 63 zones in Crimea and the occupied parts of the Zaporizhzhia, Kherson, Donetsk and Luhansk regions. It shows which towns went dark, how deep the outages went and how long they took to come back. A separate layer carries testimony from 36 local Telegram channels — kept out of the detector so it can serve as an independent check. Of those 63 zones, 48 hold enough light to measure — the rest are too dark for the method to say anything honest.
SatPredictor
Forecast of geometric-visibility windows for the Russian Rassvet constellation over Ukraine — 72 hours ahead, plus an archive of past windows. Computed from live orbital elements from Celestrak, for any single point or as a grid across the country. The elevation mask is adjustable and every forecast carries its own confidence rating.
Russian strike claims
A compilation of statements by the Russian MoD and pro-war Russian media in which a specific Ukrainian enterprise is directly named as a strike target. Who was named, when and how many times — with a source link for every claim. These are claimed targets, not confirmed hits: names get attached to strikes on a district, and the same plants are declared "destroyed" several times over.
What these tools do NOT do
The satellite cannot see who threw the switch. It sees that the lights went out — and darkness looks the same whether it came from a strike, a scheduled outage or cloud cover. Black Marble data arrives with an 8–12-day lag, so the Map of Darkness is retrospective: it is about what has already happened, not about right now.
A satellite visibility window is geometry, not a link. It means a spacecraft is above the horizon as seen from a given point. Whether the payload is switched on, where the gateways are, what the capacity is and whom the beams are allocated to — all of that is unknown, and none of it is claimed here.