The Null came back six weeks after the first probe. Not three months. Not the historical average. Six weeks.
It came at 4:22 AM on a Sunday, because hostile dimensional entities apparently didn't observe weekends.
Nox was asleep. Yara was awake. She was always awake when it mattered, as if her Compiler perception had developed a passive alert function that woke her when the architecture changed.
She didn't call Nox this time. She called the team channel.
By 4:35 AM, five Compiler users were in the mapping lab. Nox, Yara, Park Somi, Chen Wei, Han Jae. Sera arrived at 4:38, fully dressed, notebook in hand, because Sera had apparently developed the ability to go from sleep to operational in under three minutes.
The probe was different from the first.
Stronger. More directed. Instead of gentle pressure across the entire Fracture boundary, the Null was pressing hard at a single point. A focused attack against the weakest section of the dimensional membrane.
The adaptive defense module activated. The evolutionary mutation engine generated its first real-world response variation. A defensive pattern that hadn't existed before the probe began -- born in the moment of contact, shaped by the Null's attack vector, optimized through three rounds of evolutionary iteration in less than a second.
The defense held. The Null's pressure pushed against the bridge's security layer and met resistance that adapted to its approach in real-time.
"It's reading the adaptive response," Yara said. Her Compiler was wide open, tracking the probe's behavior with the sharpest perception in the room. "The Null is analyzing our defense as fast as we're adapting it. It's trying to find the frequency."
"The frequency changes with each iteration," Nox said. "That's the point. Evolutionary defense."
"But the Null is evolving too. Its probe is shifting. Each time our defense adapts, the probe adapts to the adaptation. We're in an arms race. Real-time."
The monitoring display showed the contest. Blue lines for the bridge's defense. Red lines for the Null's probe. Each line shifting, adapting, trying to outpace the other. An evolutionary race at the speed of dimensional computation.
"The Null's iteration speed is faster than ours," Park Somi said. She was reading the data with her engineer's precision. "Our adaptive defense iterates once per second. The Null is iterating three times per second."
"It's bigger," Nox said. "More computational resources. More processing power. It can out-iterate us."
"Then we lose."
"Not necessarily. Iteration speed isn't everything. The quality of each iteration matters. Our mutations are co-designed by the Plane's central intelligence -- they're optimized for this specific dimensional architecture. The Null's iterations are generic. It's applying a consumption protocol that works across dimensions. Effective but not specialized."
"Specialized beats general?"
"In the short term. In the long term, if the Null adapts its general approach to our specific defenses, we lose the specialization advantage."
The probe lasted twenty-seven minutes. Twenty-seven minutes of real-time evolutionary combat between the bridge's defense and the Null's attack. Twenty-seven minutes where the monitoring display showed two intelligences trying to out-think each other through rapid-cycle mutation and selection.
At twenty-seven minutes, the Null withdrew. Not because the defense won. Because the probe had gathered enough data. The Null was a strategic entity. It probed to learn. Twenty-seven minutes of evolutionary engagement had taught it more about the bridge's defense than a hundred gentle tests would have.
"Data harvest," Nox said. "It got what it came for."
"What did it learn?" Sera asked.
"That our defense adapts. That the adaptation is co-authored by the Spirit Plane's intelligence. That the bridge's security layer is more sophisticated than the Plane's native defense system was." He stared at the display. The red lines were gone. The blue lines settled back to baseline. "And that we're not fast enough to out-iterate it."
The room was quiet. Five Compiler users and one researcher processing the fact that the Null was faster than them.
"We need more iterations per second," Yara said.
"We need more Compiler users," Park Somi said. "Distributed iteration. If five Compiler users can produce one iteration per second, fifty Compiler users could produce ten."
"We don't have fifty Compiler users ready for bridge-level work."
"Not yet. But the seed awakenings are still producing variants. The training program is scaling. In three months, we could have twenty variants at protocol-layer capability."
"In three months, the Null probes again. Maybe sooner."
"Then we need a force multiplier. Something that increases our iteration rate without requiring more users."
Nox looked at Yara. She was staring at the display with her sharp eyes tracking the residual data from the probe's withdrawal. Her fingers were still.
"You have an idea," he said.
"The Plane's mutation engine generates variations through evolutionary algorithms. We're using it as a black box. Input threat data, output defense variation. But the engine is running on the Plane's standard processing allocation. It's not prioritized."
"The Plane allocates resources based on threat level. During the probe, threat level was elevated. The engine received priority allocation."
"But not maximum. The Plane was also running the lease protocol, the seed filter, the monitoring framework, and everything else. During a probe, the defense should get everything. Full allocation. Maximum processing."
"That would degrade other systems. The lease protocol would stutter."
"For twenty-seven minutes. Existing Weavers lose skill efficiency for twenty-seven minutes while the bridge defense runs at maximum. That's the trade-off."
Nox opened his mouth. Closed it. Yara had just proposed deliberately causing the same kind of system degradation that her own exploit had caused -- a temporary reduction in global Weaver capability to boost bridge defense.
But the context was different. A skill stutter during a probe versus a skill stutter during normal operations. One was a tactical sacrifice during combat. The other was a systemic failure during peace.
"We'd need to warn people," he said. "If the lease protocol degrades during a probe, every Weaver on the planet needs to know it's temporary and intentional."
"Build an alert system. When a probe is detected, global notification goes out. 'Bridge defense activated. Skill efficiency temporarily reduced. Estimated duration: X minutes.' People can take cover."
"You're describing a global air raid siren for dimensional probes."
"I'm describing sensible emergency communications. The Null is going to keep probing. People deserve to know when it's happening."
---
The global alert system took three days to build. Straightforward code. The bridge's monitoring function detected probe activity, triggered a notification through the bounded protocol's communication channels, and transmitted to every connected Weaver's Spirit Core.
The notification manifested as a subtle vibration -- a pulse in the Spirit Core that every Weaver could feel. Not painful. Not disorienting. Just present. A tap on the shoulder from the dimensional architecture saying: something is happening. Be aware.
Testing the system required a global deployment. Every Weaver on the planet would feel the test pulse. Nox coordinated with the Accord council. Announcements went out through every nation's Weaver command structure.
At noon on a Thursday, Nox triggered the test.
Seven million Spirit Cores vibrated simultaneously. A gentle pulse. Three seconds. Then silence.
Reports flooded in. "Felt it." "Like a phone buzzing in my chest." "My kid thought it was funny." "My cat hissed." Mrs. Zhou in Meishan province reported that her steamed buns rose faster during the pulse and wanted to know if that could be made permanent.
"System works," Sera said.
"System works. Now we need the priority allocation protocol."
---
The priority allocation protocol was harder. It required the Spirit Plane's central intelligence to agree to a temporary resource shift during probes -- diverting processing power from the lease protocol to the defense engine. The lease protocol wouldn't fail. It would operate at reduced efficiency. Eighty percent instead of a hundred. For the duration of the probe.
Nox negotiated with the Plane in the Root Directory.
```
PROPOSAL: entity(nox_renn) → process(root)
— subject: resource_allocation during null_probe events
— request: temporary priority shift from lease_protocol to defense_engine
— reduction: lease_protocol operates at 80% during probes
— duration: probe_duration only
— justification: defense iteration rate must exceed null iteration rate
```
The Plane's response was immediate.
```
RESPONSE: process(root) → entity(nox_renn)
— assessment: AGREED
— condition: automatic_reversion when probe ends
— condition: notification_system must be active
— note: process(root) has been running defense at insufficient allocation for 200 years. this correction is overdue.
```
The note made Nox pause. The Plane had been under-resourcing its own defense for two hundred years because the Fracture's energy drain left insufficient processing capacity. The compatibility patch had improved things, but the defense allocation was still constrained by two centuries of institutional underfunding.
The priority protocol wasn't a sacrifice. It was a correction. Giving the defense system the resources it had always needed and never had.
"Overdue," Nox said to the Plane. "I'm sorry it took this long."
The impression: patience. The Plane had been patient for millions of years. It could be patient a little longer. But the patience was running low. The Null was getting bolder. The probes were getting stronger. The window between probes was shrinking.
Something was changing on the other side of the scar. The Null was preparing for something larger than a probe.
---
Nox deployed the priority protocol and briefed the Accord council on the probe results. The briefing was clinical. Data-driven. Deliberately calm. The kind of briefing designed to inform without inciting panic.
After the briefing, in the hallway, Chunwei stopped him.
"How long?" the general asked.
"The Plane's assessment is that the Null is preparing for a larger action. Not an exploratory probe. An actual attempt. Timeline unclear."
"Your assessment?"
"The probing interval is shortening. First probe was on historical schedule. Second probe was six weeks later instead of three months. If the interval keeps shrinking, the next probe is in three weeks. And if the Null decides the next action is more than a probe..."
"Then the bridge needs to be complete."
"The bridge's security layer is operational. The adaptive defense is functional. But the bridge itself -- the structural gateway that transforms the Fracture into a managed interface -- is only forty percent complete. The communication layer. The access control system. The foundation. They're deployed. But the dimensional gateway component isn't built yet."
"What's needed?"
"The gateway requires code that operates in both dimensions simultaneously. Human-side architecture and Plane-side architecture, linked by the translation layer we built during the pair programming session. Nobody has ever built a bi-dimensional gateway before."
"Then do it."
"It's not that simple. The human-side architecture requires a physical anchor. Not just code in the Spirit Plane. Physical infrastructure in our dimension that interfaces with the Plane's architecture through the Fracture point."
"What kind of physical infrastructure?"
"I don't know yet. The Plane has suggested that the physical anchor should be built at the Fracture's geographic center. The point in the physical world that corresponds to the Fracture's location in the Spirit Plane."
"Where is that?"
"Zone Null's surface projection. The deepest point of the Spirit Plane maps to a physical location in the real world. Tong's research identified it fifty years ago."
Chunwei's expression didn't change. But his hands, behind his back, tightened.
"The geographic center of the Fracture," he said. "The surface projection of Zone Null."
"Yes."
"That's in Daxia. Forty kilometers northwest of the Institute."
"Yes."
"Build your anchor there. I'll clear the area and provide military support." He turned. Walked three steps. Stopped. "Renn. How many people need to be at the anchor site for the bridge completion?"
"The full Compiler team. Plus the Spirit Plane's central intelligence working through the bounded protocol. Plus a military defense perimeter in case the Null probes during construction."
"I'll have the site ready in a week."
He walked away. Military stride. Straight back. The general returning to work because the work was larger than it had been yesterday and the clock was shorter and the enemy was closer.
Nox watched him go and thought about the word "bridge." About what it meant to build a structure that existed in two dimensions simultaneously. About the fact that the hardest part wasn't the code or the architecture or the defense.
The hardest part was the trust. The decision to open a door in a wall when you knew something was on the other side trying to get in.
But a locked door was stronger than a crumbling wall. And the wall was already crumbling.
He went back to the mapping lab. Started designing the gateway.
