Everything was working.
The thought itself made Sera nervous — the scientific equivalent of a superstition, the part of her brain that had learned through thirty-seven failed experiments and two explosions that the moment you believed you were succeeding was the moment the universe corrected your optimism. But the data didn't lie, and the data was good.
Kang's second crystal growth attempt was on hour thirty-six. The furnace hummed on the secondary workbench, maintaining the precise temperature and mana field density that coaxed crystal lattice out of doped substrate. Through the observation window, Sera could see the crystal forming — a structure the size of a marble, pale violet, growing in slow accretion layers that caught the furnace's internal light and scattered it into fragments of color.
Resonance frequency at hour thirty-six: 3.61 terahertz. Climbing. The tertiary compound's higher resonance profile was doing exactly what Kang had predicted — pushing the crystal's natural frequency past the threshold that raw fluid doping hadn't reached.
"The growth rate suggests maturation at hour fifty," Kang said during his second scheduled access window. He was sitting at the monitoring station, checking the furnace's output data with the calm precision of a man who'd been growing crystals since Sera was in diapers. "Current trajectory puts final resonance at approximately 3.72 terahertz. Within your recipe's tolerance."
3.72. Two hundredths of a terahertz above the minimum. A tuning fork that might actually ring like a bell.
"Structural integrity?" Sera asked.
"The lattice is clean. No defects visible at current magnification. The compound doping is more uniform than the raw fluid — the hybrid molecular architecture integrates with the quartz substrate more smoothly. This crystal is going to be better than the first attempt in every measurable way."
Better. Not just adequate — better. The kind of word Sera had learned to distrust because it invited the universe to disagree.
"How much more growth time does it need?"
"Fourteen hours. Don't touch the furnace." He gave her the look — the one he reserved for moments when her impatience threatened to override her judgment. "The crystal is growing correctly. Let it grow."
---
Hwang's call came at 1600 on day forty-three.
"I have a contact. Professor Liu Wenxian, formerly of Tsinghua University's Department of Mana Sciences. He left academia six years ago and now operates a private supply network for dungeon biologics across Southeast Asia."
"A gray-market dealer."
"A scientist who found academic funding insufficient for his research ambitions. He specializes in Chinese dungeon materials — moss extracts, mineral compounds, biological specimens from gates that the Chinese government doesn't officially acknowledge." Hwang's voice carried the particular neutrality she used when describing activities that were technically illegal and strategically necessary. "He can supply twenty milliliters of yeongcho distillate."
"When?"
"Seven days. He has existing stock — the moss is cultivated in a controlled environment outside Chengdu, not harvested wild. The quality is reportedly consistent."
"Price?"
"Four hundred million won."
Sera's hand stopped over the tablet. ₩400 million. For twenty milliliters of moss extract. The military's annual supply budget for the entire B4 lab was ₩200 million — she was asking for double that in a single purchase.
"That includes the risk premium for violating Chinese export restrictions," Hwang continued. "Liu's network charges for discretion as well as product."
"Can the budget cover it?"
"I've reallocated funds from two other programs. The purchase is approved — I signed off this morning. The transfer is being processed through a third-party channel that Liu specified."
Approved. Already signed. Hwang had made the decision before calling Sera, which meant the colonel had calculated the cost against the value of the hack's potential and decided ₩400 million was acceptable. That calculation revealed more about Hwang's assessment of Sera's work than any briefing or performance review.
"The delivery will be handled through a diplomatic pouch from our embassy in Beijing," Hwang said. "Standard intelligence courier protocol. The yeongcho arrives at the facility in seven days. I'll have it hand-delivered to your lab."
"Thank you, Colonel."
A pause. Hwang didn't respond to gratitude the way normal people did — no "you're welcome," no acknowledgment. She processed it and moved on.
"The third Mugyeong mission. I've scheduled it for day fifty-two. Twelve days from now. The security board has approved a modified protocol — deeper penetration, extended time window, additional personnel."
"How many additional?"
"Twelve total. Eight soldiers, two field medics, you, and Sergeant Park. The largest entry team we've deployed."
Twelve people inside a living organism that had already demonstrated the capacity for strategic restructuring and cross-dimensional communication. Sera's combat potions could protect them from biological contact, help them breathe in mana-saturated atmosphere, and carve an exit if the gate tried to trap them. But twelve people was twelve bodies to protect, twelve sets of lungs to keep filtering, twelve targets for an organism that was smarter than they expected.
"I've distributed the combat potions to the team leads," Sera said. "Barrier coatings, breathing solutions, and extraction compounds. I'll brief the full team on usage protocols before the mission."
"Good. One more thing." Hwang's voice shifted — the subtle modulation that preceded information she was obligated to share but would have preferred not to. "The budget reallocation for the yeongcho purchase required a justification report. The report describes the ingredient as 'necessary for advanced potion research with strategic implications.' It's been distributed to the finance oversight committee."
"Which means more people know I'm spending ₩400 million on an ingredient."
"Which means the finance oversight committee — six members, four of whom have connections to external organizations — now knows that your research requires rare Chinese dungeon materials. I've minimized the details. But the existence of the purchase is on record."
Another leak vector. Another set of eyes on her work. Sera was starting to understand that in a military bureaucracy, every approval was also a disclosure, every authorization was also a vulnerability. The more resources she consumed, the more people had to know about it, and the more people who knew, the more likely someone would talk.
"Is there a way to do this without a paper trail?"
"Not within legal channels."
"And outside legal channels?"
"That question is better left unasked." Hwang hung up.
---
Sera spent the evening in the lab, doing what she hadn't done in weeks: planning.
Not reacting. Not scrambling from crisis to experiment to security breach to gate mission. Planning. Sitting at her workbench with the tablet, mapping the path from where she was to where she needed to be, with all the intermediate steps visible for the first time.
The pieces were converging.
The crystal — growing in the furnace, twelve hours from maturation, projected resonance at 3.72 terahertz. If it worked, she'd have the key component for the ability-code potion's proof of concept.
The yeongcho — in transit, seven days out, the rare stabilizing agent that would allow the potion to interact with System code without degrading. If it arrived intact.
The intermediate compounds — two of three complete, synthesized during the thirteen-hour window when the System couldn't see what she was building. The third still needed more tertiary compound, but she had four milliliters remaining, which was enough.
The recipe — mapped in her memory, committed to organic storage during the [Brew] cascade in the gate. The probability tree was complex but navigable. Each step was achievable with the materials she had or was acquiring.
For the first time since the System had modified [Brew] on day four, Sera could see the complete path from here to the hack. Not the divine-class recipe itself — that required ingredients and knowledge she was years from possessing. But the hack. The dissolution of the System's modification. The reopening of the pathway that [Brew] had been designed to access.
If the crystal worked. If the yeongcho arrived. If the intermediate compounds combined correctly. If the proof of concept succeeded.
If. If. If.
She wrote the plan on the tablet. Detailed, precise, each step annotated with material requirements, timing, and risk assessment. The kind of plan she'd never made before because she'd never had enough pieces in play to require one.
*Proof of Concept Timeline:*
*Day 44: Crystal maturation complete. Resonance verification.*
*Day 45-46: Third intermediate compound synthesis (requires 2ml tertiary compound).*
*Day 50: Yeongcho delivery. Quality verification.*
*Day 51: Ability-code potion assembly. Combine crystal, intermediates, yeongcho, and base medium.*
*Day 52: Third Mugyeong mission (parallel track — acquire real divine-resonance crystal for full hack).*
*Day 53-55: Proof of concept test. Apply ability-code potion to own [Brew] ability. Verify interaction with System modification.*
Eight days from crystal to proof of concept. Twelve days from now to the Mugyeong mission. The two tracks converging on the same week — the lab work and the field work, the synthetic crystal and the natural one, the approximation and the real thing.
It was ambitious. It was risky. It was the most organized thing she'd ever attempted, which made her deeply suspicious of herself.
---
Shin was reorganizing the Tier 2 storage when Sera found her at 2100. The reorganization wasn't necessary — the storage was fine, efficiently arranged, exactly as Shin had set it up two weeks ago. But Shin reorganized things when she was thinking, the way Sera paced and Min-su checked exits.
"Can I ask you something?" Sera said.
Shin's hands paused on a container of refined dungeon mineral. "Yes."
"The rat's behavioral patterns. You've been logging them for two weeks. Have they changed?"
"Yes." Shin set the container down. Pulled up her encrypted tablet. "The ninety-minute enclosure testing cycle has stopped. As of three days ago, the rat no longer tests the cage walls."
"It's given up?"
"I don't think so. It's redirected. Instead of testing the enclosure, it's spending that time observing the lab. Watching us work. Watching the crystal furnace specifically — it tracks the crystal growth through the observation window. Its mana emissions spike every time the furnace readings update."
"It's monitoring the crystal."
"It's monitoring *you* monitoring the crystal. The pattern is: you check the furnace data, the rat's emissions spike. You step away, they normalize. It's mirroring your attention. Tracking what you care about."
Sera looked at the containment enclosure. The rat was sitting near the observation-window side of its cage — the side closest to the crystal growth furnace. Its dark eyes were fixed on the furnace. Its crystalline structures pulsed faintly, a slow rhythm that Sera had learned to associate with focused attention.
The rat wasn't just watching the crystal grow. It was invested. It had stopped trying to escape the cage and started trying to understand the experiment. Redirecting its intelligence from self-liberation to collaborative observation.
It wanted the crystal to work.
Because the crystal was a step toward the hack. And the hack was a step toward the divine-class recipe. And the divine-class recipe was what the gate — the organism that the rat was slowly becoming part of — wanted Sera to brew.
The rat was the gate's eyes in this lab. And the gate was watching its investment mature.
"Shin. When the third Mugyeong mission happens, I need you in the lab, monitoring the rat. If the rat's behavior changes during the mission — if its emissions spike, if it starts testing the cage again, if it does anything that suggests the gate is responding to our presence inside it — I need to know in real time."
"I'll set up a direct comm link."
"And Shin."
"Yes?"
"If the rat tries to escape during the mission, don't try to stop it."
Shin blinked. "Why not?"
"Because it's A-rank, it can turn invisible, and it's smarter than most of the people on this base. If it decides to leave, the containment won't hold it and you won't stop it. I'd rather you be safe than try to cage something that doesn't want to be caged."
"What if it leaves and doesn't come back?"
"Then we've lost our best test subject and our closest link to the gate's biology." Sera looked at the rat. It looked back. The exchange lasted three seconds — long enough for something to be communicated, short enough that Sera couldn't identify what. "But I don't think it'll leave. It's invested too. It wants to see what happens next."
"How do you know?"
"Because it stopped trying to escape when I started succeeding."
---
At 0200 on day forty-four, Sera sat alone in the lab and let herself feel it.
Hope. The specific, fragile, dangerous kind that came from having a plan that might actually work. She'd been operating on adrenaline and obsession for forty-four days, driven by the twin engines of the approaching god and the System's interference. There hadn't been room for hope — only urgency, calculation, and the dogged refusal to stop.
But now the crystal was growing. The yeongcho was coming. The intermediates were ready. The recipe was mapped. The mission was scheduled. For the first time, the future contained a path she could follow instead of walls she had to break through.
She sat on the lab floor with Beaker in her lap — the cat had been sleeping on classified documents again and had reluctantly relocated to the warmth of her legs — and listened to the lab's nighttime sounds. The furnace humming. The ventilation cycling. The soft chime of the rat's crystals as it shifted in its cage.
The rat was singing again. The five-note pattern it had developed, played on the cage bars with its crystalline foreleg. But the melody had evolved — six notes now, maybe seven, the sequence longer and more complex, the tones clearer and more intentional. A composition in progress. A work being refined.
Sera closed her eyes and listened.
The music was small. Delicate. The kind of sound that couldn't fill a room but could fill a silence. The rat played its cage bars like a tiny instrument, and the notes drifted through the lab, and for a few minutes, the basement of a military research facility became a concert hall for an audience of one scientist, one cat, and one bodyguard who was standing in his corner with his eyes closed and his breathing slow and his hand, for once, nowhere near his weapon.
Sera didn't write anything on the tablet. She didn't need to. The moment was its own record.
In the furnace, the crystal grew. On the spectrometer readout, the resonance frequency climbed: 3.65. 3.67. Approaching threshold like a hand reaching for a door.
Six days until the yeongcho arrived. Eight days until the proof of concept. Twelve days until the third Mugyeong mission.
For the first time in forty-four days, Sera allowed herself to believe it might work.
The thought tasted like something she'd almost forgotten. Not quite joy. Not quite relief. Something closer to the moment before a reaction — all the ingredients measured, all the conditions met, the only thing remaining the act of combination.
The potential before the result.
She held it carefully, the way she held volatile compounds. Aware that it could detonate. Aware that holding it was worth the risk.
Beaker purred. The rat sang. The crystal grew.
And somewhere very far away, still approaching, the god continued its advance — indifferent to hope, indifferent to music, indifferent to the small and fragile plans of an alchemist in a basement who thought she might finally be getting it right.
