Rowan's network map arrived Sunday night, thirty-seven pages. He'd been building it since Friday and had apparently decided somewhere around Saturday to stop treating it as a preliminary analysis and start treating it as a forensic reconstruction.
The intermediary who'd fed bad intel to the port contact had a name: Sven Holler. Forty-one years old, pre-Awakening logistics background, part of the eastern port district's informal economy since before the gates started appearing. He moved supplies through legitimate and less-than-legitimate channels with the efficiency of someone who'd been doing it long enough to stop thinking about the distinction.
Holler wasn't compromised in the sense of being an Association asset. The picture was both simpler and more concerning.
He worked for whoever paid him correctly. In the past two months, a second client had been paying him correctly — more correctly than the baseline rate, which meant more correctly than the black market operators Holler had previously serviced exclusively. The second client had started receiving Holler's information flow two months ago, which was approximately six weeks after Kael had begun using the port contact.
Not targeted at the beginning. The second client was buying broad access — a sweep of the port zone's black market information network, including the resource cache locations and the identities of the people using them. Kael was a subset of a larger acquisition.
"Who's the second client," Kael said, reading the map at Rowan's table Sunday evening.
"Three layers of separation before I hit a name." Rowan was across the room, building the supplemental data on a second screen. "I hit it this afternoon. The name at the end of the chain is registered as a consultant in the Association's vendor network." He turned from the screen. "His name is Aldric Fenner. He's forty-nine, former hunter, B-rank combat background. He retired from active hunting six years ago and moved into what his Association registration calls 'talent development consulting.'"
"Meaning."
"Meaning he identifies promising young awakeners and connects them with resource networks, training opportunities, and guild development pathways." Rowan turned fully to face him. "He's one of three consultants in Ravenscrest who operate in this space officially. The other two are Association employees. Fenner is independent."
"And he's also buying black market information."
"He's buying talent intelligence. He wants to know who's moving resources in the early post-Awakening period, which tells him who the high-potential awakeners are before they surface in the public registry." He looked at the name on the screen. "The port contact is one node in a much larger scan. Fenner is building a talent map of the first-year cohort."
Kael sat with this. Not a targeted operation against him specifically. A broad sweep that had caught him as part of the data. Fenner probably didn't know Kael's specific identity yet — he had the port contact's resource movement data, which showed a single off-registry buyer running a consistent acquisition pattern. He knew someone was there. He didn't necessarily know who.
"The Association's mandatory enrollment," Kael said. "Fenner had access to the advanced cohort list."
"The Association's vendor network gives him access to non-confidential enrollment data, yes." A pause. "His broad acquisition of talent intelligence before the enrollment, combined with the enrollment list afterward — if he correlates the resource movement pattern with the advanced cohort members—"
"He narrows the candidate set significantly."
"To approximately the top eight based on capability assessment." Rowan looked at the map. "And you're the strongest candidate in that set based on the public assessment data."
"How long until he makes the correlation."
"Unknown. He's running this as a background operation, not a targeted investigation. The correlation requires him to actively compare the resource pattern data against the enrollment roster." He paused. "If he doesn't do that cross-reference, he never narrows it. If he does, he knows someone is you within a week."
Kael looked at the map. Aldric Fenner. A talent consultant with legitimate Association access running an information sweep that had accidentally touched his network.
"What does he want with the talent identification."
"The obvious answer is clients. If he finds high-potential awakeners before the guilds do, he can broker early development deals — connecting them with guild sponsorship, resource access, training networks. That's his stated business." Rowan paused. "The less obvious answer is that a talent sweep this broad and this early suggests he's working for someone who wants the map for strategic reasons rather than commercial ones."
"Who."
"I don't know yet. The three layers of client separation before Fenner suggest someone who doesn't want a direct link to the acquisition." He turned back to his screen. "I'll keep building."
"Build the cross-reference timeline. When would he realistically do the correlation."
"If he's running the data comparison manually — two to three weeks. If he has an automated system — the data he bought is already structured, so the comparison could run as a background process. In that case, he might already have it." He looked at the screen. "I'll try to determine which."
"Do that." Kael stood. "The circuit work."
"It's 2130."
"I know."
He ran it until 2300. The circuits were clean — the architecture was holding and improving. Three weeks to D-rank, approximately. He'd need the alternative source for the channel-density item.
---
The first mandatory training session was Monday at 0900.
The Association's training facility for the advanced cohort was in the main building's basement level — a space that had been purpose-built for mana-intensive work, the floor and walls layered with dampening material that kept ambient mana density at a stable level regardless of what was happening inside it. Better equipment than he'd had access to. He noted that without sentiment.
Fourteen people at the 0900 session. The advanced cohort had been split across three time slots; his slot aligned with Fenra Ahm and two others he recognized from the assessment. Marcus Thorne was not in the 0900 slot.
The training facilitator was a woman named Castellan — a former A-rank hunter, the kind of A-rank that came from twenty years of consistent work rather than a single dramatic achievement. She'd assessed the class lists before the session and had structured the morning accordingly.
"You're going to spend most of this session in baseline documentation," she said, to the room. "I need to know what you can do before I can build a curriculum for what you can't. Individual assessments, fifteen minutes each. If you finish early, use the conditioning room until the next person's done." She looked at the list. "Ashford. You're first."
His fifteen minutes produced the specific expression she used for results that required note-making — careful, calibrating, the face of someone updating a model.
"Your technique execution is atypical for your channel density," she said, after the session.
"I've focused on technique integration rather than raw output."
"That's not standard for this stage." She looked at the readout. "The junction efficiency is at a level I associate with training programs three to four years more developed than four months." She looked at him. "What's your training methodology."
"Circuit work focused on channel architecture rather than mana output. Integration of advanced technique fragments at a level my channels can partially execute." He said this because it was true and because he'd told Rowan the same thing and Rowan had begun documenting it and at some point it would be on record anyway. "Partial execution at too-high a level still produces channel shaping effects. The channels adapt to the attempt."
She looked at him for a moment with an expression he recognized. It was the expression people used when encountering something that didn't fit an existing framework and they were deciding whether to push back on the framework or log the exception.
"I'd like to document your specific technique work in detail," she said.
"After D-rank. The methodology is still developing."
She nodded slowly. Made a note. "You're going to be at D-rank within the month."
"That's the projection."
"At D-rank with this architecture—" She looked at the readout again. "You're going to be C-rank within three months. The architecture will support it." She turned to the next person. "Go to the conditioning room."
He spent thirty-eight minutes in the conditioning room while the other sessions ran. The room had equipment at a level above what was commercially available — training array calibrated for precise mana output measurement, circuit-assistance tools that could amplify specific aspects of the architectural work. He used the array conservatively, staying within the range of what was explainable by the methodology he'd described.
At 1015, the 0900 session's class list was complete and Castellan called a group briefing. The curriculum structure: two sessions per week on skill development, one session on dungeon theory and team dynamics. The dungeon theory sessions would include group deployment in Association-catalogued dungeons on a bi-weekly schedule.
She looked at the class list. "The mentorship program assignments are being finalized this week. You'll be matched with a first-year awakener outside the advanced cohort based on your class type and development profile. The program is voluntary for you, mandatory assessment for your mentees." She looked up. "The assignments come through Thursday."
---
Thursday, the assignments arrived at 0817.
ROWAN: *Mentorship program. Your assigned mentees: three names. Marcus Thorne is the first. The other two are standard cohort members with mana-support class types. The matching algorithm apparently combined class affinity and what they're calling "unusual development profile" as a pairing criterion.*
Kael read the message. Then the formal assignment notice from the Association, which listed the three names, the program structure — one session per week, minimum sixty minutes, logged in the Association's tracking system — and the first session date: Monday next week, 1400.
He sent Rowan: *Cross-reference Fenner's timeline against the mentorship assignment.*
ROWAN: *Already running it. If Fenner has the enrollment correlation and the mentorship assignments both, he has a direct link between the anonymous resource buyer and a named cohort member. The mentorship assignments are in the non-confidential vendor network by Association policy — they're shared with approved program partners.* A pause. *Fenner is an approved program partner.*
Which meant if Fenner had done the cross-reference at all, he now had Kael's name.
Kael put the phone down and thought about this for a moment.
Aldric Fenner, talent consultant, former B-rank hunter, buying black market talent intelligence. He'd identified someone moving resources efficiently in the post-Awakening period. He'd correlated that pattern to the advanced cohort. He'd gotten the mentorship assignments and now had a specific name.
What he'd do with the name depended on what he wanted.
If he was building a commercial talent roster, he'd reach out. Make contact, propose an arrangement, try to broker Kael's guild development. That was benign, if inconvenient.
If he was working for someone with strategic interests in the first-year cohort's composition — someone who wanted to know who the early standouts were before they became publicly visible — then Fenner's next move was less predictable.
He didn't know which it was yet.
He'd wait. Let Fenner make the next move. See what it was.
---
Monday, the first mentorship session.
The Association had set aside a small meeting room for the program — eight chairs, a table, a whiteboard. He arrived at 1352 and found Marcus Thorne already there. That was twice now — Marcus arrived before schedule. The reflex of someone who used early arrival to establish control over an unfamiliar environment.
The other two mentees arrived at 1359 and 1402 respectively. Brief introductions — a girl named Priya with a mana-storage class that was producing unusual density at an early stage, and a quiet man named Aden who had a class designated Shield-Form that he described with the terseness of someone who'd been explaining it to people who didn't understand it for four months.
Kael ran the session the way he'd run the orientation group exercise — asking questions rather than delivering answers, identifying where each person's understanding had gaps and what kind of information would actually fill the gap. An hour in, Priya had a clearer model of her density accumulation problem, Aden had a framework for his Shield-Form's deployment timing issue, and Marcus had said almost nothing.
He'd been watching.
Kael looked at him directly at the ninety-minute mark. "You've had the ability for two weeks. What's developed."
Marcus considered the question. "The passive reading has stabilized. I can priority-sort more efficiently — I'm not receiving everything at full intensity anymore. I can push things to background." He looked at his hands. "The active component — what I did in the dungeon assessment — I've been practicing on dungeon creatures in the public training runs. Lower-grade creatures." He paused. "It scales differently than I expected."
"Meaning."
"In the assessment, I disrupted the Stoneback's motor function for three seconds by pushing into its channel structure. I expected the disruption time to increase with practice." He looked at his hands. "It does increase. But the effect changes. It's not just motor disruption. The longer I maintain channel contact, the more I can—" He stopped. Started again. "Influence isn't exactly the right word. Suggest. The creature starts making decisions that favor where I want it to go rather than where it was going."
The room was quiet.
Priya and Aden were looking at Marcus with the specific expression that appeared when someone described an ability that was larger than they'd anticipated.
"That's control," Aden said.
"At low levels, against simple channel structures." Marcus looked at his hands. "I don't know what it does against more complex channel architecture." He looked at Kael. "Against another awakener, for instance."
"Have you tested it."
"No. I've been — the ethical problem is obvious. Influencing another awakener's channel structure without consent is—" He stopped. "I don't have a framework for what that means. The Church doesn't have a training protocol for my class. There's no ethics curriculum because there's no curriculum at all."
"The Association's ethical guidelines for mana-affecting abilities cover the territory," Kael said. "Section four of the registered awakener code. Non-consensual channel interference falls under restricted ability use — same category as certain combat-class abilities that affect nervous system function."
"You memorized the awakener code."
"I read it."
Marcus looked at him with the expression he'd used in the orientation room — the one that said he was filing something away.
"You've been reading a lot of things," Marcus said.
"I learn better from documents than from people."
"That doesn't match how you run this session." He said it without hostility. Just noting.
Priya and Aden had gone back to their own work, apparently deciding the Marcus-and-Kael thread didn't need their input.
"The ethics framework for your ability," Kael said. "Non-consensual channel interference is restricted. Consensual use in training or medical contexts is not." He looked at Marcus. "If you want to test the ability against a complex channel structure, there are ways to do that within the code."
"You're suggesting—"
"I'm describing the regulatory framework."
Marcus looked at him for a long moment. Then: "You're going to be very useful or very inconvenient, and I can't tell yet which one."
"Both, probably."
Marcus almost smiled. Not quite. "Yes," he said. "Probably."
The session ended at 1502. He logged the session notes in the Association tracker — the formal record that program compliance required. Brief, accurate, institutional. The kind of documentation that told evaluators the program was functioning without telling them anything they could use strategically.
He was in the transit system by 1515 when his phone buzzed.
A number he didn't recognize. The prefix was local — Ravenscrest, not Association-flagged, not in his contact list.
He answered.
"Mr. Ashford." A man's voice. Measured, practiced warmth. The warmth of someone who'd learned to sound like they were glad to talk to you without being glad to talk to you. "My name is Aldric Fenner. I've been following your development with some interest and I'd like to have a conversation about what your next few months might look like."
He kept his voice even. "Where did you get this number."
"The talent development network is smaller than most people realize." A brief pause — not hesitation, presentation. "I know that sounds opaque. I'm happy to be transparent about my methods when we sit down." The warmth again, precisely deployed. "I think you'll find I can offer things that the Association's standard development track can't."
Kael looked at the transit window. The city going past in the particular morning light that turned everything slightly more ordinary than it was.
"When," he said.
"Whenever is convenient. I'm flexible." A beat. "Shall we say Wednesday."
He said yes.
