The second meeting with Dr. Mira Salim happened forty-eight hours after the first text.
Not the warung this time. She had named a different location a public library branch in Depok, third floor, reference section, the kind of place that was populated enough to provide cover and quiet enough to allow a conversation. He arrived early again. She arrived exactly on time, which he had come to understand was her way not early, which would imply anxiety, and not late, which would imply carelessness.
She sat down without greeting him and placed a single printed photograph on the table.
He looked at it.
It was a surveillance photo the quality of a long-lens camera shot, slightly grainy, taken from an elevated position looking down at a street-level café. Two people sat at an outdoor table. One was a man in his fifties in a business suit whose face was partially obscured. The other was a woman in her early thirties with compact self-contained posture and a bag more functional than decorative.
The woman was Sera Vandermeer.
The man, despite the partial obscuring of his face, was recognizable by his suit, his build, and the specific way he held his coffee glass, which Rivan had observed twice a month for the past three months.
Prof. Dr. Handoko Wirawan.
Rivan looked at the photograph for a long time.
"When was this taken?" he asked.
"Three weeks ago. Singapore. Raffles Place." Dr. Salim's voice was level the tone of someone delivering information they have verified and re-verified before presenting it. "Four days before Sera Vandermeer messaged your secondary account for the first time."
The silence in the reference section had the specific quality of spaces where people are supposed to be quiet heavy, pressurized, holding more than its nominal weight.
"You've been following her," Rivan said.
"Since she appeared in the Kripto Underground ID group. I recognized the username from my research the Vandermeer name appeared twice in cross-referenced data connected to Meridian Analytics, which has three client entities in the Soerjo Capital network." She paused. "I followed her to Singapore. I was there when she met Handoko."
"You went to Singapore."
"I told you I had one trusted contact currently in Zurich." A slight pause. "I lied about the Zurich part. He was in Singapore. He took the photograph."
Rivan set the photograph down carefully. He looked at the water stain on the library table old, circular, the kind left by a glass set down thousands of times in the same spot. He looked at it for approximately four seconds, which was the time it took to complete the recalibration that the photograph required.
"She's working for them," he said.
"I don't know," Dr. Salim said.
He looked at her. "You have a photograph of her meeting Handoko Wirawan four days before she contacted me. That seems-"
"Suggestive. Yes. But not conclusive." She leaned forward slightly. "Consider the other possibility. Handoko met with her. Not the other way around. Who initiated that meeting I don't know. It's possible she went to him. It's equally possible he summoned her. And those two scenarios have entirely different implications."
"If she went to him-"
"Then she may have been trying to access the network from the top rather than the bottom. A risky strategy but not an irrational one for someone who has been investigating independently for fourteen months and has hit the same walls I have." She held his gaze.
"And if he summoned her-"
"Then she's been deployed against me specifically. And everything she showed me two days ago was a controlled intelligence feed."
"Yes."
They sat with it.
"The Singapore data she showed me," Rivan said. "The circuit analysis, the unnamed node, the @v_salim_research username
-"
"The username pointing to me," Dr. Salim said, with the flat, unembellished accuracy of someone naming the thing in the room. "Yes. I noticed that you would find it. I've been waiting to see what you would do with it."
"It's your username."
"It was my username. On a research forum I used under a pseudonym in 2017 and 2018. I stopped using it after the Seoul paper was retracted I became more careful about my digital footprint." She looked at him steadily. "Sera Vandermeer either found it through genuine forensic research, which would mean she is as good an analyst as she appears. Or it was given to her fed into her dataset specifically to make her appear credible to you while simultaneously casting doubt on me at the critical moment when you were deciding how much to trust each of us."
Rivan looked at her.
"You understand what this means," he said slowly. "Either you're telling me Sera is compromised and I shouldn't trust her. Or Sera is legitimate and the username she found is real which means you have a connection to the Singapore network that you haven't disclosed."
"Yes," Dr. Salim said. "That is exactly the situation."
"And you're telling me this why?"
"Because the correct response to an impossible choice between two explanations is not to pick one and commit. It is to keep both active, operate with maximum caution toward both, and find the evidence that resolves the ambiguity." She paused. "I'm telling you because if I were the compromised party, I would not be handing you the framework that allows you to suspect me equally."
"Unless handing me that framework is itself the move," Rivan said.
Dr. Salim looked at him with something that was not quite a smile but occupied the same general territory. "You've gotten sharper in the past month."
"I've had reasons to."
He left the library at 4 PM and walked.
He did not take his usual route back to the boarding house. He walked the longer way, through the campus perimeter, past the economics faculty building where Handoko Wirawan's office sat on the second floor with the window that caught the afternoon light. The window was lit. Someone was in there.
He did not stop walking.
He was building two models in his head simultaneously the way he had learned to run parallel analyses in his original life, keeping competing hypotheses alive without collapsing them prematurely into a single narrative. One model had Sera as an agent of The Architects, deployed to infiltrate his investigation and either extract information or redirect it. The other had Dr. Salim as the compromised party, using the photograph and the forum username to preemptively discredit Sera before Sera could share something Dr. Salim didn't want shared.
Both models fit the available evidence.
Both models had gaps.
The gap in the Sera-is-compromised model: the data she had showed him about the unnamed Singapore node was genuinely useful, genuinely detailed, and genuinely consistent with everything he had independently observed. Controlled intelligence feeds were designed to feel accurate but they were also expensive to produce, because every accurate piece of intel you fed an adversary was information they could use. Why give him that much?
The gap in the Salim-is-compromised model: she had told him Handoko was The Curator, which was either true in which case she had handed him the most operationally significant piece of intelligence he possessed or false, in which case she had given him a target that was wrong, which would eventually be verifiable, at which point her credibility would collapse completely. A long con with a built-in expiry date.
He stopped walking.
Stood on the pavement outside the university's east gate while motorcycles moved past him in the warm April afternoon, and looked at nothing in particular while something assembled itself in the back of his mind with the slow, inevitable quality of a pattern completing its final iteration.
The GHOST Protocol appeared not with its usual clinical readout, but with something he had never seen before. The overlay was the same translucent layer it always was, but the text that assembled itself did not look like its standard outputs. It looked like something the system was generating in real time, uncertain, the way a person sounds when they are thinking out loud:
[ GHOST PROTOCOL ]
⚠ Processing conflict: Two incompatible threat models active.
Standard resolution: insufficient data.
Non-standard observation: —
Both Dr. Salim and Sera Vandermeer have shown you
things that were true.
Both have withheld things that were also true.
This is not the behavior of an enemy.
This is the behavior of someone who is also afraid.
Revised classification for both entities:
NOT ENEMY. NOT ALLY.
SAME SITUATION.
They are doing to you what you are doing to them.
Everyone in this room is lying about the parts
they are most afraid to lose.
Recommendation: Stop trying to determine who to trust.
Start determining what everyone is afraid of.
The answer to that question is the same for all of you.
Rivan read it twice.
Everyone in this room is lying about the parts they are most afraid to lose.
He thought about Dr. Salim four years of research, a journalist named Dewi Hartanto who no longer had a byline, a trusted contact she had initially lied about being in Zurich. A woman who had built an entire investigation in silence because the alternative was to become one of the names on her own list of disappeared researchers.
He thought about Sera fourteen months of working alone, a supervisor who had told her to move on, a firm with three clients connected to the network she was investigating. A woman who had flown to Jakarta to talk to a stranger with a silent Telegram account because she had run out of safe options.
He thought about Laras four months of dataset building, a thesis framework constructed as cover, a question she had asked him in the coffee place that had the precision of someone who already knew the answer: Do you know the name?
He thought about himself. Twelve years of failure compressed into nine weeks of action, a system in his vision he still did not understand, a skripsi supervisor who might be the architect of everything that had destroyed him, and a secret he had held so tightly for so long that the shape of it had become invisible to him the thing he was most afraid to lose.
The answer to that question is the same for all of you.
What were all of them afraid of?
Not exposure. Not failure. Not even death, the disappeared researchers had not stopped others from investigating.
They were afraid of finding out that the thing they were fighting was larger than the fight they had been having.
That Soerjo Capital was not the top of the structure. That The Architects were not the ceiling. That whatever was above Handoko Wirawan whoever had designed a system sophisticated enough to predict crypto markets with institutional precision in 2019 was something none of their combined datasets had yet touched.
He started walking again.
He had four people now, himself, Dr. Salim, Laras, Sera, each with a different piece and a different angle. He had capital that was growing toward the level he needed. He had a timeline he understood better than anyone else alive.
And he had, for the first time, a question that was the right size for the problem.
He pulled out his phone. Opened a new message, not to Dr. Salim, not to Sera, but to Laras, who had been quiet for two weeks and whose silence he had been reading as patience rather than absence.
He typed: "I need the four of us in the same room. Can you arrange a neutral location? Not the campus. Somewhere none of us have a prior association with. R"
He put the phone in his pocket and walked toward the boarding house in the warm April light.
Behind him, on the second floor of the economics faculty building, the light in Handoko Wirawan's window was still on.
Three hours later, his phone showed three new messages.
From Laras: "Tuesday. I know a place. I'll send the address. L"
From Dr. Salim: "Agreed. But I want to know who the fourth person is before I walk in. M"
From Sera, which meant she had a channel to him that he had not given her, which was itself a piece of information:
"I already know about the other two. Laras Andini and the researcher who was in Depok last month. I've known since before I messaged you. I've been waiting for you to put it together yourself, because people trust conclusions they reach independently more than conclusions they're handed. SV"
"Also, the photograph Dr. Salim showed you. Ask her who took it. The answer will tell you something important about her that she has not told you yet. SV"
"Tuesday works. SV"
Rivan stared at the three messages for a long time.
Then he opened his notebook to a fresh page, wrote Tuesday at the top, and began to prepare.