The National Bureau's city division occupied the top four floors of a government-grade building in the Central District. It did not try to be impressive. It had been doing its job for sixteen years and had the specific authority that came from being the official record — the database that every faction, practitioner, and awakened in the city had registered in at some point, whether they'd wanted to or not.
He came at two PM, as scheduled.
The building's lobby had the institutional feel common to official facilities — high-tier security wards on the entry points, the faint qi-dampening field that all National Bureau facilities maintained at the entry level. Not hostile, just professional: the Bureau needed to be able to assess anyone who walked in without the ambient cultivations of strong practitioners contaminating the assessment equipment. The lobby's dampening field would reduce his probability field's passive range from thirty-odd meters to about three. He walked through it and felt the scope reduce.
The entry clerk at the desk ran a standard card scan and directed him to the elevator. Floor thirty-one.
Han Weiwei's assistant, a precise B-rank practitioner with the particular manner of someone who had worked for a very direct superior long enough to absorb some of the directness, met him at the elevator. Not introduced herself — identified him from the cultivation field and said: "Mr. Chen. This way." Led him to the thirty-first floor without small talk.
Adjacent to the data archive center. He had expected this — the administrative logic of keeping the primary analysis office next to the primary data source was obvious to anyone who thought about it, but not everyone thought about it.
The office was smaller than Lin Zhengyue's but more densely occupied. Not cluttered — organized, with the specific organization of someone who worked primarily by document and spatial arrangement of documents. Three workstations, two research surfaces, one meeting area near the window. The window faced northeast, which meant the city's financial district was visible and the Central District's cultivation field saturation was lower than it would be facing south or west.
She was at the secondary research surface when he came in, not her main desk. The main desk had one item on it: his file. The public version, plus one additional folder he hadn't seen before.
She turned.
---
She was smaller than her Bureau ID photo suggested. This was becoming a pattern with women whose file photos didn't convey scale. She was thirty-one and looked closer to her age than most high-rank practitioners did — the SS-rank candidate's physical enhancement hadn't been running long enough to change her baseline presentation significantly. But the speed of her initial assessment was SS-trajectory speed: she'd covered the distance from the research surface to the meeting area and processed his full cultivation signature in the same motion.
The spatial manipulation ability was invisible at rest. He could feel the ambient asymmetry of it as a slight irregularity in the local probability field — the way available positions in the room had a slight lean toward her, as if space itself had a mild preference for her needs. Not a perceivable effect to anyone without a probability field. He registered it clearly.
He thought: she is going to be very good at this in six months, and very careful about it now, and she has been doing research for fourteen months without anyone explaining to her what she was looking at.
She assessed him in the time it took to cross the office. She was fast. The assessment was complete before they shook hands.
"Mr. Chen," she said. "The intake runs about fifteen minutes. I'll do that first."
"All right."
---
The formal intake was exactly that — fifteen minutes, the standard classification registration for new S-rank practitioners. But even within the formal structure, she was doing something else. The intake form was a guide, not a script — she went off it three times to ask clarifying questions that the form didn't prompt for. Name, registration number, cultivation method designation. He listed it as Hybrid Foundation System, nonstandard architecture, which was what the Vanguard had registered. Ability type: Field Effect, Probability. He explained the designation briefly and she wrote it down without prompting him further, which meant she had already looked up the designation and found nothing and had decided to ask about it separately.
Tier confirmation, current organizational affiliation: Celestial Vanguard, Senior Operative, two days ago.
The clarifying questions:
First: "The hybrid foundation — is the multi-element architecture stable at S-rank, or are there ongoing recalibration requirements?" He told her the fourth anchor had been the longest phase of the breakthrough and that post-completion settling ran for two weeks. She wrote this down with the specific precision of someone filing it not for the form but for her own analysis.
Second: "The zero-assessment registration from sixteen months ago — the Bureau's original assessment used the standard protocol for that period. Have you had the original assessment record reviewed?" He told her the original assessment was accurate by the protocol's measurement criteria. He had nothing the standard framework measured at the time. She looked at him when he said *at the time* and did not follow up on it during the intake.
Third: "The Celestial Vanguard formalized your operative status two days ago. Prior to that, you were contractor-classified for over a year. Were there any operational incidents during the contractor period that should be captured in the S-rank registration record?" He told her nothing in his operational history required formal Bureau capture. She wrote this down without comment.
She worked through the form with the speed of someone who had done this two hundred times and whose attention was on the next part of the conversation.
At fourteen minutes, she put the form down.
"Done with the formal intake." She closed the official folder and opened the other one — the one with the three tabs. She looked at him directly. "Field Effect, Probability. That designation has come through this office exactly once in sixteen years. Today."
"I know."
"It's not in the standard ability registry."
"The ability is external-source, not awakening-event distribution," he said. "The registry was built on awakening-event ability types. The nonstandard types don't have registry entries because the standard framework didn't anticipate them."
She looked at him.
"External source," she said. She repeated it with the weight of someone who had been building toward a specific word and just had it confirmed.
"Yes."
She opened the second tab — the one he was interested in, the one she'd put the most work into. She spread three pages on the desk: his registration timeline, a progression curve chart she had built herself from secondary data, and a probability distribution graph that he had not seen before. All three pages were annotated in the margins, some of the annotations in multiple colors, the revision marks of someone who had been returning to the same documents repeatedly over months and finding new things to add.
"Your progression curve," she said. "Zero-assessment, sixteen months ago. F-rank registered nine weeks later. D-rank, six weeks after that. C-rank at month six. B-rank at month nine. A-rank at month twelve." She looked at him. "S-rank six days ago."
"Yes."
"No practitioner in the Bureau's sixteen-year record has progressed from zero-assessment to S-rank in under four years," she said. "The fastest previous case took three years and two months, and that was a mutually reinforcing dual-ability awakened with above-average natural cultivation affinity." She paused. "You registered zero. You have no natural cultivation affinity on record. Your progression is faster than the fastest natural case by a factor of approximately three."
He said nothing.
"The probability distribution graph," she said, touching the third page. "This is the distribution of awakening-event ability types in the city's 3% allocation. The standard model predicts this distribution based on latent potential in the population. The actual distribution matches the model at 94.7%." She looked at him. "The remaining 5.3% don't match the model. They cluster in specific ability types and demographic segments in ways the standard model can't explain."
He looked at the graph.
The distribution shape was familiar — he had looked at distribution models in the system's information packages on two occasions. The standard bell curve, the 3% awakening-event allocation sitting well above the tail, the various ability types represented as discrete clusters within that allocation. The graph looked the same as the standard model except for the upper-left cluster.
That cluster didn't belong in the standard model. There was no mechanism in the standard framework's theoretical structure that should have produced the high-latent-potential, zero-at-assessment anomaly. The standard model said high latent potential correlated with high probability of awakening-event ability allocation. The upper-left cluster said: some of the highest latent potential cases got nothing.
The 5.3% deviation was visible as that cluster — the high-latent-potential, zero-at-assessment cases. The category he had been in.
"You've been tracking this for eight months," he said.
"Fourteen months," she said. "I started when I was in the analytical division, before I became senior director. I brought the research with me when I transferred." She closed the folder. "I have six identified cases of practitioners who registered zero-assessment and subsequently developed cultivation or ability capacities that don't match the latent-potential activation model. You are one of the six."
"I know," he said.
She looked at him.
"You knew I had this research."
"I inferred it," he said. "The message you sent was specific enough to indicate you had a working hypothesis that my file confirmed."
She held his gaze. "What do you know about the other five?"
He thought about Tang Mingxiu's evening conversation. About the name she'd mentioned. About the city's three identified cases — one being himself, one likely being the person Tang Mingxiu worked with, one being the western district practitioner Tang Mingxiu had described.
"Two by contact," he said. "One by description from a secondary source."
She went still.
"You've identified three in this city," she said.
"Two confirmed and one probable," he said. "You've identified six, with geographic spread beyond this city."
"Three in this city." She said it slowly, processing the arithmetic. "Three in other municipalities." A pause. "If there are six you know of and three in one metropolitan area — the city isn't a random concentration."
"No," he said. "The city's density of mechanism cases may be specific. The mechanisms may have been deployed to locations where recipients would eventually interact."
She absorbed this. "You've been building the same picture I've been building."
"From inside it," he said. "You've been building it from the distribution data. I've been building it from lived experience with one of the mechanisms."
"Mechanisms," she said. The plural was intentional and she heard it.
"The six you've found are six separate mechanisms," he said. "The same design architecture, probably, but individual instances. Each one is specific to its recipient."
"How specific?" she said.
"The tasks the mechanism generates are calibrated to the recipient's professional capabilities and personal circumstances," he said. "My mechanism uses my computer science background, my mathematical training, my professional contacts. The tasks are achievable by me specifically. They would not be the same tasks for a different recipient."
She looked at him.
"The mechanism knows you," she said.
"The mechanism was built for me," he said. "Or built for a profile that I happen to fill."
"The selection criteria," she said. She was writing now, the pen moving quickly across her notes. "The mechanism selects for the recipient first, then builds the task architecture around what that recipient can do." She looked at him. "Which means the selection happened before the deployment."
"Before the Awakening Event, in all likelihood," he said.
She stopped writing.
She was quiet for a moment. The assessment behind her eyes was moving faster than the conversation — she was processing faster than she was speaking, running the implications of each confirmation.
"The design architecture," she said. "The external source."
"Yes."
"Someone built these," she said.
"Yes."
"Before the Awakening Event," she said, "or after?"
He thought about the Unmarked Package's single line of text. *The Awakening Event had an architect.*
"The evidence suggests before," he said. "The mechanisms were deployed within days of the event's completion. That suggests they were prepared in advance."
She absorbed this for four seconds — he counted.
"The Awakening Event was designed," she said. Not a question. She had been moving toward this conclusion for fourteen months and she had just received the confirmation that pushed it from hypothesis to working theory. "The distribution. The 3% selection. The compensatory mechanisms for the cases the distribution missed." She looked at him steadily. "Someone designed all of it."
"The evidence points that direction," he said.
"Do you know who?"
"No," he said. "I have data points but not a picture. The Awakening Event predates my active research by sixteen years."
She looked at the folder on her desk. Then back at him.
"I want to work this together," she said. Directly, without ceremony. "My distribution data and your mechanism data. The combined picture is more complete than either of us has. The research gets to the answer faster."
He thought about the system's MQ4 notation. The approach will develop naturally.
"I can share data," he said. "On a calibrated basis — some of what I know has implications I need to think carefully about before disclosure."
"Understood." She met his gaze. "I'm asking for what you can give, not everything. The research needs partial truth more than it needs silence."
"Thursday next week," he said. "I'll bring a data set."
She nodded once. Then: "The seven-province interregional report," she said. "I've been trying to get the full dataset from the national archive for six months. There's an administrative hold from the Zhao era that's still active."
"I know," he said. "I submitted a review request before I came up today. The hold should clear within forty-eight hours."
She stared at him.
"You submitted the request."
"Through proper channels. The hold should have been cleared when Director Zhao retired. It wasn't. The review process will complete it."
She looked at him for a moment with an expression he was beginning to recognize as her specific version of recalibrating — when a variable she hadn't accounted for arrived and the model needed updating.
"The interregional report covers my three highest-probability external cases," she said. "If it confirms them, the global distribution picture expands significantly."
"I know," he said. "That's why I submitted the request."
She was quiet for a long moment.
"Thursday," she said. "Bring what you can."
"All right."
He stood. She stood.
"One more question," she said.
He turned.
"The probability field," she said. "Field Effect designation. You said it scales with cultivation tier. Does it have an observable limit? Any tier where the effect plateaus?"
He thought about the five levels. Level 5: reality accommodates him.
"Not visible from where I am," he said.
She looked at him. Not with alarm — with the specific focus of someone who had just been told something that fit into a larger picture they were building, and the fit was precise.
"That," she said, "is the most interesting answer anyone has given me in the last six months."
He noted the six months. Not an idle phrase — it matched the timeline of when the distribution anomaly research had entered its current phase. She had been asking this question for six months, in various forms, to practitioners who hadn't known the answer. He was the first who did.
He left the Bureau at four.
The lobby's dampening field came back over him as he walked through it. His probability field's range expanded back to its standard thirty-odd meters the moment he cleared the entry threshold. The city's ambient cultivation field registered in full resolution — the Central District's density, the specific signatures of practitioners he'd catalogued in his range.
He walked back through the afternoon light and thought about the seventeen-recipient picture and the woman who had assembled six of them from distribution data alone, working backward from outcomes to the structure that produced them. She had been doing for fourteen months what the system had built him to do from the inside.
Different angles. The same shape.
Four to five months to her SS-rank certification. The system had said the approach would develop naturally. He had expected "naturally" to mean gradually — a slow accumulation of contact and credibility before the research collaboration could begin. Instead, the first meeting had opened all the way to the designed-event hypothesis in one session.
This was the probability field operating at S-rank. Not just luck. Pattern recognition at scale — the field surfacing not just favorable outcomes but the most direct path to beneficial structure. The meeting had run the way it ran because both of them had the information to arrive there fast, and the field had created the conditions for that speed.
He would need to calibrate his expectations. S-rank wasn't just bigger. It was faster.
Four PM on a Thursday. Lin Zhengyue's building was visible from this angle of the Central District — the Lin Family headquarters two kilometers southwest, the upper floors catching the afternoon light. The section five revision session was in three hours.
He walked toward the subway and turned over what he was going to bring to the next Thursday meeting. His mechanism architecture data — LP structure, tier progression, the probability field's calibration levels at each tier. The seven-province interregional report, if it cleared in time. And the questions he couldn't answer: who the architect was, what the cultivation site held, what the convergence of all seventeen recipients was supposed to produce.
The research had started. He thought it had started in the right direction.
He took the subway south, toward the Lin Family district and the working session and the evening ahead.
