The coordination center was busy with morning activity when Marcus arrived.
Chen was at the main table, coordinating supply runs. Julie was updating the security board with overnight entity sightings. Maybe fifteen other people scattered around the room, all doing the routine work of keeping two hundred survivors fed, housed, and alive.
Marcus scanned faces, trying to figure out who'd triggered the anomaly flag. The headache from forcing the cooldown still throbbed behind his eyes-a persistent reminder of the cost.
His skill's passive detection was pinging-that same background sense that something was wrong-but it wasn't specific enough to identify a target. He needed another active scan, but his stamina was still recovering.
Chen spotted him and waved him over. "Webb. You look like death. Rough night?"
"Testing the skill upgrade. Found something weird."
"Weird how?"
Marcus lowered his voice. "My active scan flagged an anomalous player signature. In this room. Eighty-two percent confidence."
Chen's expression shifted from concerned to calculating. "When?"
"Eight minutes ago. Forced a second scan-dropped to sixty-three percent but the signal's consistent. Someone here is reading as unusual to the System."
"Unusual how? Everyone here is Displaced. We're all anomalous by default."
"Not anomalous like Displaced. Anomalous like..." Marcus struggled for the right words. "Like the System thinks someone's behavior or status doesn't match expected parameters. Like when you run a sanity check and one data point doesn't fit the model."
Julie had drifted over, listening. "Can you identify them?"
"Not without another scan. Cooldown's almost reset. But if I scan in here, whoever it is will know I'm looking."
"Or," Chen said quietly, "you're detecting yourself. You just accepted a telemetry-flagged skill upgrade. Maybe you're the anomaly."
Marcus hadn't considered that. It made uncomfortable sense-he'd just had code installed that let the System track his every observation. Of course that would mark him as unusual.
But the scan had shown his own position separately. The anomalous signature had been distinct from his location.
"I don't think it's me," Marcus said. "The signature was twenty meters from my position. Someone else."
His cooldown timer hit zero.
Marcus triggered an active scan without warning.
The debugging overlay snapped into place, and for 8.3 seconds he could see everyone in the room rendered as entity signatures with floating metadata. Chen: [Level 5 Coordinator | Status: Normal | Behavior: Management]. Julie: [Level 3 Logician | Status: Normal | Behavior: Analysis]. Others scattered around the room, all showing up as [Status: Normal].
Except one.
Near the supply board, a woman Marcus vaguely recognized-Level 3 Crafter, mid-thirties, practical clothes, short dark hair-showed up with different tags.
[Level 3 Crafter | Status: MONITORED | Behavior: Observation | Flag: System Interest]
The scan ended and Marcus's vision snapped back to normal.
"Her," he said, nodding toward the woman. "Near the supply board. System's flagged her for something."
Chen followed his gaze. "That's Rivera. She was on Team Seven-part of the building save yesterday. One of the first-shift leads."
"Why would the System flag her?"
"Same reason it flagged you," Julie said. "She was part of the exploit. Everyone on Team Seven participated in gaming the deprecation system. If the System's tracking who's involved in anomalous behavior, they'd all be marked."
That made sense. But it didn't explain why Rivera's flag was different-why she showed up as anomalous while others read as normal.
Marcus pulled up his pattern history and checked the timeline. Rivera had been in Building 2847 from 12:00 PM to 4:00 PM yesterday-the exact window when the activity index had climbed from 0.03 to 0.14 and the deprecation flag was rescinded.
She'd been there when the exploit worked. When the System's maintenance routine failed. When they'd proven that structural deprecation could be reversed.
The System's not just watching me, Marcus realized. It's watching everyone who participated. Tracking them. Monitoring for repeat behavior or pattern development.
"We need to check the others," Marcus said. "Everyone on Team Seven. See if they're all flagged."
"Why?" Chen asked.
"Because if they are, the System's not just interested in the exploit-it's interested in the people who execute it. It's building a profile of exploit behavior. Learning who in the cluster is dangerous enough to monitor."
"And if they're not flagged?"
"Then the System's selection criteria are more specific, and we need to figure out what makes Rivera different."
Chen pulled out his phone and sent a text. Thirty seconds later, Patterson-the Level 4 Defender who'd been on Team Seven-entered the coordination center.
Marcus triggered another scan.
Patterson showed up as [Status: MONITORED | Behavior: Security | Flag: System Interest].
Same flag. Same monitoring tag.
"They're all marked," Marcus said. "Team Seven's been flagged by the System."
Julie leaned against the table, expression thoughtful. "So participating in an exploit gets you put on a watchlist. That's... actually predictable. The System identified unusual behavior, flagged it for analysis, and tagged everyone involved for follow-up monitoring."
Unauthorized duplication: this narrative has been taken without consent. Report sightings.
"Which means," Chen said, "every time we find and use an exploit, we're increasing our visibility. The more we game the System, the more attention we draw."
"Not just attention," Marcus corrected. "Active monitoring. Everyone who helped save Building 2847 is now being tracked. The System's collecting data on them-movement patterns, resource usage, probably every action they take. It's building behavioral profiles."
"For what purpose?" Rivera had walked over, clearly aware she was being discussed. "So it can deprecate us faster next time we try something?"
Marcus shook his head. "I don't think so. If the System wanted us dead, it would just kill us. But it's not doing that-it's watching. Like it's running an experiment."
"An experiment with us as test subjects," Rivera said flatly.
"Yes. Which means we're valuable for something. The System wants data from us-probably to patch the exploit we found. It needs to understand how we did it so it can prevent future attempts."
"So we taught the System how to defend against building saves," Patterson said. "Great. That's fantastic."
"Or," Julie said slowly, "we taught it that we're capable of finding exploits. That we're threats. That we need more aggressive monitoring."
The room was quiet for a moment.
Then Chen spoke. "Webb, give me your assessment. Is this monitoring hostile or passive? Are we in immediate danger?"
Marcus thought about it. Reviewed the data. Considered patterns.
"Passive," he said finally. "If the System wanted to eliminate us, entity deployment would be aggressive-Horrors breaching perimeters, active attack formations. What we're seeing is containment. Increased presence, strategic positioning, but no direct assault. The System's treating us like a problem to study, not a threat to destroy."
"Yet," Rivera added.
"Yet," Marcus agreed. "But that gives us time. Time to find more exploits, learn more patterns, figure out how to survive this."
"While being watched," Patterson said.
"While being watched," Marcus confirmed. "That's the trade we made when we saved the building. We proved we could fight back-and we got the System's attention in return."
The council meeting that afternoon had the feel of a post-mortem.
Word of the monitoring flags had spread through the cluster. Two hundred people now knew they were being watched, and opinions varied sharply on what to do about it.
Director Vasquez had called the meeting for 2:00 PM-giving people time to process the news but not so much time that rumors could spiral into panic. The west conference room was packed. Council members. Team leads. Anyone who had a stake in the cluster's survival strategy.
Marcus stood near the front with Chen and Julie, trying to ignore the weight of attention. Last night he'd been the hero who saved a building. Today he was the person who'd confirmed they were all on the System's watchlist.
Vasquez called the room to order. "Status update. Webb discovered this morning that participants in yesterday's building save have been flagged for System monitoring. Anyone on Team Seven is now being actively tracked. Webb, explain what that means."
Marcus pulled up his documentation on a borrowed tablet and projected it on the wall screen. "The System tags certain players with a 'MONITORED' status flag. Based on my testing, this triggers enhanced observation-your actions, locations, and resource usage are tracked more precisely than baseline. It's not immediate threat, but it is elevated scrutiny."
"How many people are flagged?" someone asked from the back.
"Everyone on Team Seven-twelve people. Possibly others who participated in planning or logistics. I haven't scanned the entire cluster."
Rodriguez-the pragmatic Scout who'd questioned Marcus yesterday-raised her hand. "What's the threat model? What can the System actually do with this monitoring data?"
"Best guess: behavioral analysis. It's learning our patterns so it can predict and counter future exploits. Worst case: targeted elimination-deprecating monitored individuals first, removing the people most likely to cause problems."
The room shifted uncomfortably.
"Have we seen evidence of targeted elimination?" Vasquez asked.
"Not yet. Deprecation seems random so far-or at least, not correlated with System attention. But that could change. We're in new territory."
A man Marcus didn't recognize-Team Four lead, logistics-spoke up. "Then we should stop. Pull back on exploit discovery. Don't give the System more reasons to pay attention."
"That's assuming attention is bad," Chen countered. "We don't know what 'behavioral analysis' means long-term. It could be routine. It could be the System just logging data without taking action."
"Routine? The System doesn't do routine. It does optimization." The logistics lead looked around the room. "Every time we game one of its mechanics, it learns. Every time we find an exploit, it gets smarter. We're teaching it how to defend against us."
"Or," Julie said, "we're learning its rules faster than it can patch them. That's how exploit cycles work-discovery, exploitation, patching. The question is whether we can leverage the window between discovery and patch."
"And if we can't? If the System adapts faster than we do?"
"Then we lose," Marcus said bluntly. "But we're losing anyway. Deprecation's happening regardless of whether we fight back. At least this way we're learning. At least we have a chance."
Vasquez tapped the table-her characteristic considering gesture. "Webb, Chen-your recommendation?"
Chen spoke first. "Continue exploit discovery, but with protocols. Small tests. Controlled conditions. Document everything. Assume we're visible and plan for System response. We don't stop looking for advantages-we just get smarter about testing them."
"Webb?"
Marcus thought about it. About the telemetry in his skill. About Team Seven's monitoring flags. About the System watching, learning, adapting.
"Same," he said. "We can't afford to stop. But we need to be strategic. Not every exploit is worth the attention cost. We prioritize high-value discoveries-things that save lives or buy time. We avoid testing things that might trigger aggressive response without meaningful benefit."
"Define 'aggressive response,'" Rodriguez said.
"Entity swarms. Accelerated deprecation. Targeted elimination. Anything that suggests the System's moved from passive monitoring to active threat suppression." Marcus paused. "If we see that, we back off immediately. Attention is manageable. Hostility isn't."
Patel, the Synthesist, raised her hand. "What about your skill telemetry? Should we be worried about that specifically?"
"Yes and no. The System's tracking my Anomaly Detection usage, which means it knows what I'm finding. But it also means I have better tools to find things worth knowing. Net assessment: the benefit outweighs the cost, but only if I'm strategic about what I scan."
"Which brings us to the operational question," Vasquez said. "How do we use this information? Do we expand building saves to other structures? Do we test the individual deprecation hypothesis with Webb's timeline? Do we keep pushing?"
The room was quiet for a moment.
Then Marcus spoke. "I think we push. But carefully. We expand building saves to high-priority residential structures-places where deprecation would hurt the most. We document the process, refine the protocols, make it repeatable. And we test my individual deprecation timeline-because if we can reverse individual flags, that changes everything."
"And the monitoring?" the logistics lead asked. "What do we do about the System watching us?"
"We accept it," Marcus said. "We're already visible. The question isn't how to hide-it's how to use the visibility window before it closes. We find exploits, we use them, and we learn faster than the System can adapt. That's our advantage. That's how we win."
"If we can win," someone muttered.
"If we can win," Marcus agreed. "But the alternative is waiting to be deprecated without fighting back. I don't like those odds better."
Vasquez looked around the room, reading faces, gauging consensus. "Vote. Continue exploit discovery under controlled protocols, accepting System attention as an operational cost. Show of hands."
Hands went up. Not universal-fourteen out of twenty-three present-but enough.
"Motion carries. Chen, coordinate building save expansion. Webb, you're primary on testing and documentation. Anyone on Team Seven who wants off the monitoring list can request reassignment-no penalty, no judgment. Questions?"
"Yeah." Santos, the young Runner who'd been on Team Seven, raised his hand. "What happens if the System decides we're too much trouble? If monitoring turns into something worse?"
Vasquez met his eyes. "Then we adapt. Same as we have been. We're survivors. That's what we do."
"Survivors with target tags," Santos muttered.
"Survivors who just proved we can fight back," Vasquez corrected. "That's worth something. Dismissed."
The council filed out, conversations spinning up immediately-debate, concern, cautious optimism. Marcus stayed near the front with Chen and Julie.
"That went better than expected," Chen said.
"Half the room thinks we're committing suicide by attention-seeking," Marcus said.
"The other half thinks we're buying time. That's consensus in a crisis." Chen pulled out his phone. "I'm organizing expansion logistics. You focus on documentation and your own timeline. Forty-eight hours left on your deprecation, right?"
Marcus checked his status screen. "Forty-six hours, twelve minutes."
"Then let's make them count."

