The man in the jacket had a name.
It took me nine days to find it, and finding it required spending money I would rather have kept, calling in an information contact I had been saving for something more significant, and sitting through a two-hour conversation with a source in Akihabara who believed that the value of his information was directly proportional to how long he took to deliver it.
His name was **Arata Fumio**.
Fifty-one. Former ward-level investigator, -war municipal corruption division. His division had been dissolved in the restructuring — not because corruption had stopped being a problem but because the reconstruction government needed the personnel budget elsewhere and corruption, in a city still pulling bodies from collapsed buildings, felt like a secondary concern to most people.
Arata had not agreed with that assessment.
He had not left the work. He had just stopped being paid for it.
I sat with that for a while.
---
The profile built itself once I had the name.
He had been a mid-tier investigator with a decent clearance rate, two internal commendations, and one formal reprimand for pursuing a case past the point where his superiors had decided it wasn't worth pursuing. The reprimand was logged but hadn't ended his career, which meant he was valuable enough to keep but inconvenient enough that they wanted him to know there were limits.
He had a one-bedroom apartment in Nerima. He did occasional contract work for a private insurance firm — fraud assessment, legacy claim verification, the kind of work that an ex-investigator could do quietly and that nobody glamorous was competing for.
No current institutional affiliation. No registered PI license, which meant anything he was doing now was either informal or careful enough to stay invisible.
He was asking about the Kanda office informally.
Which meant either nobody had sent him, or whoever had sent him was careful enough not to leave a visible connection.
I assessed both options and their implications.
Option one: independent curiosity, thread pulled from a previous job, low-level pattern recognition. Manageable. He was one person with no institutional backing, and one person with no institutional backing was a pressure I could redirect.
Option two: someone with enough sophistication to use an unaffiliated former investigator as a probe. That was more expensive and more deliberate, and it implied a principal who had already decided the Kanda office was worth investigating, which meant the principal had a reason.
I didn't know which option it was.
So I decided to find out directly.
---
I spent two days building a context.
Not a cover identity — I didn't need one. I needed a plausible reason for a chance encounter in a location Arata frequented, with enough ambient legitimacy that the encounter wouldn't register as constructed.
He ate lunch on Tuesdays at a teishoku restaurant in Nerima that had survived the war because its owner had been stubborn about the structural reinforcement and had turned out to be right. He arrived at 12:15, sat at the counter, ordered the same thing with minor variations, and left by 1:00.
I arrived at 12:30 on a Tuesday with a reconstruction contractor's site assessment folder and the specific kind of tired that field workers carry in their posture. Not performed — I had been awake since 5:00 AM reviewing Hatsue's latest phishing infrastructure reports and the tiredness was real. I just let it show.
I sat two seats down from him. Ordered. Did not look at him.
He looked at me twice in the first five minutes. Occupational habit — read the room, identify variables.
I let him read me.
What he would see: mid-twenties, construction-adjacent paperwork, unremarkable clothing, no obvious affiliations. Someone who belonged in a workman's lunch counter in Nerima. Someone with nothing to notice.
At 12:48, I asked the counter staff about a ward office address, out loud, with the mild irritation of someone who had been given contradictory directions.
Arata answered before the counter staff could.
"The ward annex moved. It's on the east side of the station now, near the temporary housing blocks."
I thanked him. He nodded and went back to his food.
I let forty seconds pass.
"You know the ward well," I said. Conversational. Not pressing.
"Used to work there," he said.
"Reconstruction staff?"
"Before that."
He wasn't offering more. That was fine. I wasn't asking for more.
We ate in parallel silence for four minutes, which is a long time in a counter restaurant and felt natural rather than awkward because I had been careful about the quality of the silence. Not pointed. Not waiting. Just two people who had run out of things to say and were comfortable with that.
He left at 12:58.
I stayed and finished my food and thought about what I had learned.
---
What I had learned was small but precise.
He was observant without being aggressive. He had answered my question reflexively — a former civil servant's instinct to be useful — and then pulled back when the conversation could have continued. He wasn't looking for connection. He was comfortable with distance.
That was the profile of someone who operated alone by preference, not by circumstance.
Option two was less likely. A principal using him as a probe would want him more socially engaged, more positioned to pull information conversationally. His instinct was to withdraw, not to extend.
Probably option one.
Probably.
I added a third column to his file: *Assess: independent actor, motivated by pattern recognition, low institutional backing. Risk level: low-moderate. Action: monitor. Do not escalate unless contact repeats.*
Then I noted underneath: *Probably is not definitely. Do not close the file.*
---
The team dynamic had shifted in a way I was still measuring.
Riku's conversation slip had done something subtle. Not to his performance — he had been exactly on-task since our conversation, precise and contained in the way people are when they're being careful to demonstrate something. But it had changed the way the others positioned themselves.
Hatsue had become slightly more formal in briefings. Nori had started arriving two minutes early. Daigo had said nothing different but was watching me in the way he watched targets — with the specific patient attention of someone building a read.
I had created a small fear event without intending to.
That was not entirely bad. A functional amount of uncertainty about consequences was useful in a team structure. People who had no concept of what leadership would do under pressure became sloppy in ways that couldn't be corrected quickly.
But fear had a half-life, and past its useful range it became resentment, and resentment in a team that operated through close coordination was a structural problem.
I recalibrated.
The next briefing I ran loose — walked through the Shinkō Data consultancy numbers with visible satisfaction, asked Hatsue a genuine question about her methodology that I didn't already know the answer to, let Nori explain a technical variation he'd developed unprompted and responded to it with the attention it deserved.
Not warmth, exactly. More like respect made visible.
By the end of the hour the room felt different. Calibrated back to functional.
I was aware that I was managing them the way I managed everything else — by understanding what state I needed them in and applying the appropriate input.
I was also aware that this was, in some technical sense, a form of what I did professionally.
I didn't examine that for long.
---
Shinkō Data's second client came through a referral from the first.
The concrete supplier had mentioned us to a logistics coordinator who was trying to map functional cold-chain routes for medical supply distribution in the northern wards. Different problem, same underlying structure: information that existed somewhere in the reconstruction bureaucracy but wasn't accessible through normal channels because the normal channels were buried in their own backlogs.
Hatsue spent three days on it.
The final assessment was clean, specific, and delivered in a format that looked exactly like the kind of thing a small data consultancy would produce. Professional without being polished enough to seem like something that required explanation.
We charged sixty-five thousand.
The coordinator paid, asked if we could help with a follow-on question, and referred us to two other logistics firms.
I looked at the referral chain.
One client becomes two becomes four. Not because we were selling hard. Because in a reconstruction economy, functional information was scarcer than money, and people who found a reliable source told other people who needed the same.
We had, almost by accident, built the beginning of a reputation.
I told Hatsue to register a proper service description on two reconstruction business directories. Nothing elaborate. Just visible enough to catch the right kind of search.
She asked what kind of clients I wanted.
"Mid-tier logistics. Materials procurement. Anything adjacent to ward-level contract decisions."
She understood why without me explaining. Our information pipeline was deepest in exactly those areas. We could service that client profile better than any legitimate competitor because our sourcing was incomparably better.
The legitimacy was the layer that made the sourcing invisible.
---
The Yaoyorozu file reached a threshold that week.
I had been building a timeline — her board attendance, her assessment contributions, the delay between her input and Ikeda's formal recommendations, the contract approval dates. When I laid it flat and looked at the pattern over five months, something clarified that I hadn't seen at the point-level.
The contracts she was most substantively involved in weren't random. They clustered around a specific materials category: high-tensile composites for structural reinforcement. The kind of materials that had been in critical shortage since the war, that required either expensive importation or sophisticated on-site fabrication.
Her quirk could generate those materials directly.
Which meant she wasn't just consulting on contracts for materials that needed to be procured. She was potentially producing materials that were then being logged as procured through external contractors.
That was not charity work.
That was something else entirely.
I opened a calculator.
If even a fraction of the contracts in that category had been fulfilled through direct Creation quirk application — with the procurement costs recorded as paid to contractor firms that may or may not have actually delivered anything — the financial gap between the logged expenditure and the actual production cost was enormous.
Someone was extracting that gap.
She might know. She might not.
If she knew and was participating: different problem, different approach.
If she didn't know — if she had been producing materials under the genuine belief that she was contributing to reconstruction, while someone else was logging the fictional contractor payments and extracting the difference — then she was not a participant.
She was a resource being exploited by people who understood her well enough to know she would cooperate with reconstruction work and trusted institutional frameworks enough not to ask the right questions.
I leaned back in the chair and looked at the ceiling.
The Suidobashi ventilation system hummed.
I thought about what it meant to know this.
It meant I had something on Ikeda. Potentially something on the principal above Ikeda. It meant the financial scale of what was happening around Yaoyorozu was significantly larger than her public profile suggested. It meant that approaching her — at the right moment, with the right frame — would not just be an information acquisition operation.
It would be something more complicated.
The thing underneath the operational calculus, the thing I kept in the separate column, surfaced briefly:
*She is being used by people who knew she would trust them.*
I noted it.
Closed the column.
Returned to the operational calculus.
---
I ran a smaller job that Friday — credential extraction from a private arbitration firm in Minato that had been handling reconstruction contract disputes. The target had a specific document set I needed for a Shinkō Data client assessment, and the legitimate channel for obtaining it would have required two months and a formal request that would have left our name in a paper trail I didn't want.
The job ran in eight minutes. No team deployment — just Nori remote, myself on physical, a protocol I had been developing for small-footprint operations where full team presence would be disproportionate to the target.
It worked.
Afterward I ran an assessment: how many jobs in our current portfolio were actually single or dual-person operations that I had been overstaffing because I had designed every protocol around the full team?
The answer was most of them.
I restructured the protocol tiers over the weekend.
Tier one: full team, high-value physical presence required.
Tier two: three-person, moderate complexity.
Tier three: one or two, remote or limited physical.
Most of what we were doing was tier three.
The team's time was a resource. I had been spending it imprecisely.
---
Sunday evening I walked the Shinjuku route I had walked with Puppimil months ago.
Not to meet anyone. Just to think in motion, which worked better for me than thinking in the office because the office was where I solved specific problems and Shinjuku was where I let the larger shapes surface.
The reconstruction had hit this area harder than Suidobashi. The commercial strips were partially rebuilt — new storefronts in old bones, neon signs above facades that didn't quite match. The izakayas were mostly back. The corners that had been dangerous before the war were dangerous again, which in a strange way indicated normalization.
I thought about Arata Fumio, eating his lunch alone.
I thought about Daigo taking notes on his teammates.
I thought about Yaoyorozu asking questions nobody answered, and continuing to show up anyway.
There was a specific kind of person who kept showing up to a system they were starting to distrust. Not naive. Not trapped. Something more specific: someone who believed that the right thing was still the right thing to do even if the people around them were doing it wrong, and who hadn't yet decided whether that belief was strength or liability.
I had known a few people like that.
Most of them had been hurt by it eventually.
Visit My Patreon and read upto 30+ Advanced chapters in advance to support the story
My Pat reon:ht tps://patreon.com/eyestapler
Thanks for reading