Jin's contact came through in thirty-six hours, which was faster than Rook had expected and meant the contact had prioritised the request, which meant Jin had told her something about what was actually at stake.
Rook decided not to ask. Trust, in the context of information that couldn't go through official channels, was the decision to let other people make the judgment calls within their area.
The original sensor data from K-7's 2147 intercept arrived as an encrypted package in Jin's personal storage. They decrypted it in the systems bay at 03:00, when the station was quiet, which was becoming their working pattern.
The data was more substantial than the anomaly flag had suggested.
The official flag had described a "non-standard noise pattern, possible stellar phenomenon." The raw sensor data told a different story: a structured signal, clearly encoded, thirty-seven seconds of transmission at a frequency that K-7 had been tuned to detect but that no known stellar phenomenon produced. K-7's onboard analysis had classified it as noise because its encoding didn't match any known artificial signal protocol.
It didn't match any known artificial signal protocol in 2147.
In 2187, Rook had access to forty more years of xenocommunications research — most of it theoretical, all of it classified at levels above hers, some of it accessible through Jin's considerably more creative approach to data clearance, which Rook was also not going to ask about.
She matched the encoding against the xenocommunications database.
It wasn't a match. It was something closer to a prototype — the same general structural logic as a communication protocol that had been theorised in 2173 by a researcher at Kepler Institute and then quietly classified. The 2147 signal used a simplified, lower-resolution version of the same structure.
"Whatever sent this in 2147," Rook said, "was using a protocol that humanity didn't develop until 2173."
Jin looked at the comparison on the screen. She was very still.
"Or," she said, carefully, "humanity developed the protocol in 2173 because they were working backward from the 2147 signal."
Rook looked at her.
"The Kepler research," Jin said. "Where did they get the starting data?"
Rook queried the archival record of the 2173 Kepler research. The starting data was listed as "original theoretical construct, see primary investigator notes." Primary investigator notes were classified.
She thought about what "see primary investigator notes" meant when the notes were classified at a level above a deep-space analyst's clearance.
She thought about a company that had, three times in twenty years, reclassified anomalous findings.
"I need to know who owns the 2173 Kepler research," she said.
Jin was already looking it up. "Kepler Institute is a subsidiary," she said. "Of Meridian Research Consortium." A pause. "Meridian Research Consortium is —"
"The parent company of Meridian Systems," Rook said.
They both sat with that.
Outside, the transit corridor was dark and empty and eleven million kilometres away was a point in space that something had indicated, through a dead probe and a forty-year-old signal, was going to matter in thirty-one days.
"They knew," Rook said. The words were quiet. "They received the 2147 signal. Kalephas logged it. The research consortium classified the study into it. They've been sitting on this for forty years."
"We don't know that," Jin said, carefully.
"We know the signal was received. We know the protocol research was classified. We know the parent company owns both the research and this station." Rook stopped. "We know that this morning I received a follow-up signal from the same source and the protocol tells me to log it through a chain that runs directly to the parent company."
Jin was quiet for a long moment.
"What do you want to do?" she said.
Rook looked at the coordinates. Eleven million kilometres. Thirty-one days.
"I want to put a sensor sweep on those coordinates," she said. "Passive — nothing that transmits, nothing that would register as active surveillance. I want to know if there's anything physically there. And I want to do it without triggering the station's standard observation log."
Jin looked at the sensor array schematics on her secondary screen. "I can route a passive sweep through the maintenance diagnostic channel. It won't register as a survey operation. It'll take six hours."
"Can you do it tonight?"
Jin nodded.
Rook looked at the decoded 2147 signal and then at the seven words from this morning and thought about the specific quality of decisions that didn't feel like choices so much as acknowledgements of what you were already committed to.
"One more thing," she said. "The message says *do not let them.* Assume for a moment that K-7 — or whatever's using K-7's signal architecture — knew this message would reach someone. Assume it was designed for a specific recipient."
"You," Jin said.
"Me, or someone like me. Someone in this station, working this frequency, at this time." She paused. "It knew I would be the analyst on shift. The same way the informant knew which detective would be assigned the tip."
She stopped. That was a different story — she had been thinking about a book she'd read once, and her mind had produced the reference without warning. She let it pass.
"If it knew who would receive this," she continued, "then *do not let them* was addressed to me specifically. Not as an instruction to the station, not to the company. To me."
"Which means you're already implicated," Jin said.
"Yes."
"Whether you want to be or not."
"Yes."
Jin looked at her with the expression of someone who was going to make a decision and had already made it and was only waiting for the form of words to say so. "I'll start the sensor sweep tonight," she said. "We'll see what's there."
Rook nodded. She pulled up the coordinates one more time on her screen and looked at them: the quiet specificity of numbers that mapped to an apparently empty region of space where, thirty-one days from now, something was apparently going to arrive.
*They are coming.*
She hoped the sweep showed nothing.
She was fairly certain it wasn't going to show nothing.
