The footage reached the Calloway’s office within minutes of the first posts going live.

His deputy had been awake since the first posts went live, running collection on the flagged accounts, and had assembled the preliminary file before Calloway’s car cleared the security gate. The orbital footage was flagged at the top with a single annotation: Watch this one first.

Calloway watched it standing at his desk, with his coat still on.

He watched it again. Then he sat down slowly and watched it a third time.

The skeleton filled the upper half of the frame. The Voyager — the vessel that had crossed the solar system on live broadcast — sat beside it and looked small.

The construction drones moved across the superstructure in organized patterns, hundreds of them, their movement continuous and purposeful, and the lunar surface curved at the bottom of the frame like a reminder that this was real and not rendered.

He pulled up the communication system. "Get me Reeves. Get me the NRO duty officer. And get me whoever is currently running the lunar telescope array." A pause. "All three. Simultaneously."

***

Calloway was on with Reeves within the hour.

"I need a position model," he said, before Reeves had finished his greeting. "Built from the orbital footage the staff posted this morning. I need you to extract every usable reference point — the first spacecraft’s orbital position, the sunlight angle on the skeleton, the curvature of the lunar surface in the frame, the apparent altitude, the angular relationships between the visible structures. Everything the footage gives you." He paused. "Then I need you to run those reference points against our lunar orbital geometry database and give me an estimated surface location for the base. Where on the moon is that footage being taken from? That’s the question. I need the answer with a confidence interval and I need it fast."

"You want me to reverse-engineer the base location from a video posted on LucidNet," Reeves said.

"I want you to try," Calloway said. "And I want you to assume the footage is genuine, because everything we know about these accounts says it is. Work from that assumption. Tell me where they are."

Reeves was quiet for a moment. "Give me two hours."

"You have ninety minutes."

***

Reeves called back in eighty-three with the results ready.

"I have the model," he said. "I ran it three times independently before calling you. Each pass produced the same result within a forty-kilometer variance."

"Where?"

Reeve paused fit s moment, as he considered choosing between the answer that sounded reasonable and the answer that was accurate.

"Far side," Reeves said.

The line went quiet.

"You’re certain?" Calloway asked.

"I’m certain the reference points are consistent and the model is sound," Reeves said. "The footage gives us the Voyager’s orbital geometry, the sunlight angle at the time of recording, the visible curvature of the lunar surface, and the angular separation between the two structures. When you run those variables against known lunar orbital mechanics, the position that fits them falls on the far side. Every time... I also sent the footage to two external contractors without context. Didn’t tell them what I’d found or what they were looking at. Asked for independent position estimates."

"And?"

"Both came back far side. Within thirty kilometers of my estimate."

Calloway stood up from his desk.

He walked to the window and stood there, looking at nothing, and his deputy watched him from across the room.

"Task the NRO," Calloway said finally, without turning from the window. "Full far-side coverage. Every asset with lunar far-side capability, tasked now, highest priority. I want imagery of the estimated location and the surrounding region." He paused. "And pull the telescope array. If there’s external illumination on that base — and the footage shows significant external illumination — it’s emitting light outward. The far side faces deep space, not Earth, so it won’t show up in Earth-facing surveillance. But a telescope with the right orientation should be able to catch it."

His deputy was already on the second line, tending to the task given to her.

***

The NRO results came back in forty minutes, with clean cratered surface, with no installation visible, no thermal signature and no electromagnetic anomaly of any kind.

The telescope array also returned the same result.

Calloway read both reports standing at his desk.

"Run it again," he said.

"Frank—"

"Run it again."

The second pass took twenty minutes and the results were identical.

He set the reports down.

"The stealth systems," his deputy said from across the room. "The specifications document said that ’Stealth systems will be disabled from entry into the airport’s airspace until departure.’ They disabled them for the operation window. A voluntary gesture." She paused. "They didn’t disable them for anything else."

Calloway looked at her.

"A fixed installation on the far side," she continued. "With no Earth-facing exposure and no light emission directed toward deep space rather than toward our sensors. There will be limited coverage from our orbital assets because our surveillance infrastructure was designed around Earth-facing geometry, not far-side observation." She stopped. "And stealth systems applied to the installation itself. All four factors simultaneously. We were never going to find it."

"No," Calloway said. "We weren’t." He said it flatly, without inflection. "Not because our assets failed or because our coverage had a gap that better resources would have closed. But because they assessed our full surveillance capability accurately, identified the location where all four factors converged, and built there." He paused. "From the beginning. Whatever the beginning was."

His deputy was quiet for a moment. "How long has it been there?"

"The footage gives us the shuttle fleet in the bay," Calloway said. "The base infrastructure visible in the room tours. The scale of the common areas and the residential level." He picked up the NRO report and set it back down. "None of that is built quickly. None of that is built in a year. The construction timeline implied by what we can see in the publicly posted footage is—" He stopped.

"Long," Calloway said. "Significantly longer than any timeline we had been working from. The base has been operational, staffed, and fully built for a period of time that our model of this organization does not account for." He looked at her directly. "Whatever assumption we were working from about when Nova Technologies became real — when it became capable of this — that assumption was wrong. And we don’t have a way to determine by how much."

The room was very quiet.

"Write the assessment," Calloway said. "Everything we can confirm from the footage. The position model, the three independent estimates, the NRO results, the telescope array results, the stealth capability inference. The conclusion stated clearly — not buried, not softened." He picked up his jacket. "The President needs to understand that we are not dealing with a surveillance gap. We are dealing with an entity that understood our architecture better than we understood theirs, built accordingly, and has been doing so for longer than we know."

He paused at the door.

"They didn’t hide from us," he said. "Hiding implies they needed to avoid detection. They built where we weren’t looking. There’s a difference and it matters. One is defensive. The other is something else entirely."

After saying that, Calloway left.

His deputy sat alone in the office with the NRO report on the desk in front of her and the orbital footage still running on the screen behind it.

She looked at the footage for a long moment. Then she opened a new document and began to write.