What we know about the US strikes on Khuzestan, and what we still don't

Two large explosions were heard over Omidiyeh, in Iran's oil-producing Khuzestan province, at roughly 22:24 UTC on 12 July 2026, according to the Telegram channel intelslava. Within forty minutes a second feed, rnintel, logged further strikes on the towns of Susangerd and Andimeshk, and by 23:09 UTC an attack on Hoveyzeh. A short lull followed. Then the two channels disagreed, publicly, on whether Aghajari Airport had been hit. By 00:31 UTC on 13 July, rnintel was reporting a US strike on Omidiyeh Airport itself, the same city intelslava had opened the night with.

The pattern is rolling, attributed, and thin on confirmation. It is also the cleanest real-time record of an air campaign inside Iran that, at the time of writing, no major Western wire or Iranian state outlet has formally acknowledged.

What the two channels agree on

Both intelslava and rnintel record an opening strike on Omidiyeh within minutes of each other on the evening of 12 July, and both attribute the action to US forces. Both go on to describe further strikes on towns strung along Khuzestan's southern corridor: Susangerd, Andimeshk, and Hoveyzeh, all between 23:02 and 23:09 UTC. The geography matters. These are not random provincial targets. Omidiyeh, Aghajari and the surrounding fields sit on the Iranian section of the Ahvaz-Dasht-e Azadegan oil belt, home to some of the country's most productive upstream infrastructure. Susangerd, Andimeshk and Hoveyzeh run further north, closer to the border with Iraq and to the road and rail arteries linking the petro-economy to the Persian Gulf.

A second point of agreement is procedural. Neither channel presents footage, neither names a weapon system, neither cites an official source. Both are correlating geolocated social-media chatter with prior knowledge of US force posture in the Gulf. They are consistent with each other on roughly five discrete strike locations over roughly seventy minutes. They diverge on one important location.

Where they contradict each other

At 23:51 UTC on 12 July, rnintel posted that reports of an attack on Aghajari Airport were false. Sixty seconds later, intelslava posted that a US airstrike on Aghajari Airport, in Omidiyeh, had been reported. The contradiction is direct and unresolved in the thread. By 00:31 UTC on 13 July, rnintel had filed a follow-up placing a confirmed US strike on Omidiyeh Airport, which sits in the same urban complex as the Aghajari site; whether that subsequent strike was at the same coordinates, on a different runway, or on an adjacent military node, the channel does not specify.

This kind of live-threading error is not unusual. Telegram war channels operate on seconds-long cycles and often run on second-hand accounts from local Azerbaijani or Iraqi Kurdish networks that move faster than the Iranian opposition outlets they relay. It is, however, worth treating the Aghajari claim with more caution than the surrounding items, simply because the two feeds actively disagree about it.

What the framing already does

There is a temptation, when reading a string of Telegram items in this tone, to treat the messaging itself as proof: precise timestamps, flag emojis, the verbs of war. The temptation should be resisted. Telegram is now the primary distribution layer for both Iranian dissident networks and Iranian state-aligned media; it is also a useful channel for Western open-source analysts because geolocated video tends to surface there first. The uniform present-tense, third-person framing of these posts ("US attack on", "US airstrike against") is a habit of the genre, not a sourced attribution.

A read on what is not yet in the record is, at this point, almost more useful than a read on what is. There is, as of the items in front of this publication, no Pentagon confirmation, no CENTCOM release, no Iranian Foreign Ministry statement, no Iranian Civil Aviation Authority notice on Omidiyeh or Aghajari airspace, no quoted official from Khuzestan's governorate, no image of impact craters or of munition debris that has been independently geolocated. Two Telegram channels have moved. The architecture of attribution around them has not.

What to watch over the next 24 hours

If the strikes occurred as described, the next 48 hours will be loud. Three signals will tell us what kind of campaign this is. First, an Iranian response, which under Tehran's declared doctrine of reciprocity would be expected within hours and would likely come through Iraqi Kurdish or Gulf-state proxy channels. Second, a US acknowledgement or denial. The Pentagon's silence through the overnight hours is, on past form, suggestive rather than conclusive. Third, oil-market behaviour at the next Asian open, since Khuzestan's infrastructure sits upstream of the Strait of Hormuz chokepoint and even a threat to throughput moves the curve.

What the public should hold lightly, for now, is the strike list itself. The geography is consistent. The order of events is partly consistent. The single disputed node, Aghajari Airport, sits in the same municipality as the later, more firmly logged strike on Omidiyeh Airport, which is at least suggestive that something on the airfield complex was engaged. That is a long way from certainty about what was hit, why, or under what authority. The record will firm up fast; until it does, the cleaner sentence is the older one from the trade: wait for the second source.

Desk note: Monexus is publishing this as a sourced reconstruction of two Telegram feeds, not as confirmation of the underlying events. Where the feeds disagree, we have flagged the contradiction in line; where neither feed is corroborated by an official source, we have said so. We will update the lede and the strike list as Western wires, Iranian state media, or independent OSINT publish verifiable confirmation.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/intelslava
• https://t.me/intelslava
• https://t.me/rnintel
• https://t.me/rnintel
• https://t.me/rnintel
• https://t.me/rnintel