Sera brewed eleven potions in three days.
Not the careful, measured, one-at-a-time process she'd used for the Enhancement Elixir or the Refraction compound. This was production work. Assembly-line alchemy, driven by a deadline she'd set herself: the third Mugyeong mission needed to launch within two weeks, and the entry team needed tools she hadn't built yet.
Combat potions. Defensive compounds. Things that kept people alive in environments that wanted them dead.
Day thirty-nine, she started with barriers. A topical compound — smeared on skin or armor — that created a thin mana-reactive shield against biological contact. Inside the gate, the organisms would try to touch them. Last time, the entry team had worn standard tactical gear that offered no protection against direct biological interaction. This time, each team member would carry a barrier coating that repelled organic material the way oil repelled water.
[Brew] gave her the recipe in clean, bright probability branches. The ingredients were common — mana-reactive polymer base, refined dungeon mineral dust, cave moss extract as a binding agent. She produced eight doses in four hours, each one tested on a sample of gate tissue she'd preserved from the first mission. The tissue slid off the coated surface like it was greased. Effective.
**[Protocol-Monitored item #19: Biological Barrier Compound (A-Rank). Your cooperation is appreciated.]**
Day forty. Breathing solutions. The gate's atmosphere was mana-saturated at concentrations that degraded cognitive function within thirty minutes. On the first two missions, the team had relied on standard-issue respirators with mana filters. The filters worked, but they reduced oxygen flow and limited communication. Sera brewed a mist — inhaled through a small diffuser — that lined the airways with a mana-resistant coating, allowing normal breathing in high-saturation environments for up to two hours.
Six doses. Tested on herself — she inhaled the mist and immediately felt the difference. The ambient mana in the lab, usually imperceptible, vanished from her senses like someone had closed a window against the wind. Her lungs expanded fully. Her head cleared.
**[Protocol-Monitored item #20. Your cooperation is appreciated.]**
Day forty-one. Emergency extraction compound. This one was Min-su's request — the only time he'd asked her for anything specific.
"Something for retreat," he said. Standing at her workbench, arms folded, explaining the tactical scenario with the economy of a man who measured words the way Sera measured milliliters. "If the gate closes the corridor. If the organisms cut off the exit. Need a way out."
"What kind of way out?"
"Destructive."
Sera thought about it. The gate's organisms were biological — living tissue, mana-reactive, organic. Destructive meant something that could clear a path through organic material quickly enough to allow a team of soldiers to move from deep inside the gate to the exit before the countdown reached zero.
She brewed an acid. Not a conventional acid — a mana-reactive compound that dissolved organic tissue on contact while leaving inorganic material untouched. Poured on a gate organism, it would eat through the tissue in seconds. Poured on stone, metal, or the team's tactical gear, it would do nothing.
The recipe required precise calibration. Too weak, and the acid wouldn't clear a path fast enough. Too strong, and the mana reaction would generate heat — potentially enough to ignite the gate's mana-saturated atmosphere. She tested three formulations on preserved gate tissue, measuring dissolution rate against thermal output, before finding the balance.
Four doses. Each one in a reinforced glass container with a breakable seal — snap the seal, pour the contents, run.
**[Protocol-Monitored item #21.]**
Min-su picked up one of the containers. Weighed it in his hand. Tested the seal mechanism — thumb pressure, quick snap, clean release. He nodded once, put it in his tactical vest, and returned to his corner. The most enthusiastic review she'd ever received from him.
---
Kang's crystal frequency calibration study was approved on day forty.
The approval came with conditions — Kang was the listed principal investigator, all work had to be conducted during his scheduled access windows, and the results were to be submitted to Hwang's office within seventy-two hours of completion. Standard bureaucratic guardrails that reduced a potentially open-ended research program to a three-day sprint.
Three days. To synthesize a crystal that resonated at divine-class frequencies.
Kang arrived for his first working session at 0800, carrying a case of specialized equipment he'd requisitioned from the university's materials science lab — a crystal growth furnace, precision oscillators for frequency measurement, and a sample of synthetic quartz substrate that would serve as the seed crystal.
"The approach is straightforward," he said, unpacking the furnace onto the secondary workbench that had been rebuilt after the first crystallization explosion. "We grow a standard mana crystal using your synthesis protocol — the one that produces 97% purity. Then we modify the crystal's resonance frequency by introducing a doping compound during the growth phase."
"Doping with what?"
"The rat's crystalline structures emit at a specific frequency — coherent mana field modulation at approximately 3.7 terahertz, based on the data Shin collected. That frequency is within the divine-class range, according to the System's own classification parameters."
"You're pulling the target frequency from the rat."
"The rat is the closest available source of divine-class resonance data. Its crystals resonate at the right frequency because they're derived from the Mugyeong organisms, which are themselves—"
"Connected to the same architectural framework as the System. Yes. The frequency is the same because both the gate and the System are built on the same foundation."
"Exactly. If we dope the crystal growth with a compound that introduces the 3.7 terahertz resonance into the lattice structure, we should get a crystal that rings at the right note."
"Should."
"Physics is more certain than alchemy, Sera. The crystal will resonate if the lattice geometry supports it. My uncertainty is about whether the resonance will be deep enough to serve your purposes."
The tuning fork versus the bell. A synthetic crystal might hit the right note but lack the harmonic depth of a crystal grown naturally in the gate's core, saturated with the mana density that permeated the Mugyeong organism's deepest structures. Surface resonance versus structural resonance. The difference between humming a melody and understanding the music.
"What do you need from me?" Sera asked.
"Precise emission data from the rat. Not the averaged readings from the monitoring equipment — direct measurements, as close to the crystalline structures as I can get. And one milliliter of the black fluid, as a doping agent for the crystal growth."
One milliliter. A small amount, relative to her current supply. But every milliliter was accounted for now — tracked by the System, monitored by Hwang's office, allocated through the pre-approval process. Using fluid for crystal synthesis had been approved. Using it for a crystal specifically designed to resonate at divine-class frequencies had not been approved, because the approval request described it as "frequency calibration."
Sera retrieved the fluid. Measured one milliliter with a precision pipette. Handed it to Kang. He introduced it to the crystal growth furnace's substrate chamber alongside the synthetic quartz, and the process began.
Growth time: forty-eight hours. The furnace would maintain controlled temperature and mana field density while the crystal formed around the doped substrate. In two days, they'd have either a divine-resonance crystal approximation or a very expensive piece of tinted quartz.
---
While the crystal grew, Sera addressed the second problem: ingredients she didn't have.
The ability-code potion — the hack — required the divine-resonance crystal as its primary component. But the recipe also called for several rare compounds that she hadn't sourced: a mana-conductive mineral found only in deep dungeons, a biological extract from high-rank monster tissue, and a stabilizing agent derived from a plant that grew exclusively in the mana-saturated environments of specific Asian gates.
The mineral she had — the military's supply chain included deep-dungeon mining operations that produced mana-conductive compounds. The monster tissue extract she could synthesize from the gate tissue samples in her storage. But the plant extract — a compound called *yeongcho* distillate, derived from a luminescent moss that grew in the mana-rich caves of certain Chinese dungeons — wasn't available through Korean military channels.
China didn't share dungeon resources. Not officially, not through diplomatic channels, not through the Association's international exchange programs. Chinese dungeon materials stayed in China, processed by Chinese alchemists, distributed through Chinese supply networks.
Which meant Sera needed a back channel.
She raised it with Hwang during a scheduled briefing on day forty-one. The colonel came to B4 for the first time since the security lockdown — a gesture that signaled either flexibility or surveillance, depending on Sera's mood.
"I need an ingredient from a Chinese dungeon," Sera said. "Yeongcho distillate. The moss grows in B-rank and higher dungeons in Sichuan and Yunnan provinces. Korean suppliers don't carry it."
"The Chinese don't export dungeon biologics."
"Officially. There are back channels. Hwang, you run a military intelligence division. You have contacts in every country that has gates. I need twenty milliliters of yeongcho distillate, and I need it within two weeks."
Hwang looked at her. The controlled assessment look. Sera could almost see the calculations running behind the colonel's eyes — risk of reaching out to Chinese contacts, risk of being discovered, risk of creating another international incident on top of the Japanese inquiry.
"What's it for?"
"A compound that requires ingredients I can't source domestically. The recipe is part of an ongoing research program." Vague enough to be true. Specific enough to be useless.
"I'll make inquiries," Hwang said. "No promises. Chinese contacts are... delicate."
"Your entire operation is delicate. That's never stopped you before."
Something flashed across Hwang's face — not quite a smile, not quite irritation. Something in between that Sera had learned to read as the closest Hwang came to acknowledging a compliment.
"Two weeks," Hwang said. "I'll see what's possible."
---
Shin found Sera on the lab floor at 2200, sitting with her back against the containment enclosure, staring at the ceiling.
"You're running on fumes," Shin said. She sat down next to Sera — close, but not touching. They'd developed a physical language over the past two weeks, an unspoken agreement about proximity that reflected their working relationship: close enough to hand instruments, far enough to not crowd.
"I've been running on fumes since day one. The fumes are just higher-octane now."
"You brewed eleven potions in three days. You're running a crystal growth experiment, coordinating a military mission, and trying to source illegal ingredients from China. When did you last sleep more than three hours?"
"Sleep is a luxury good. My supply chain doesn't support luxury."
"Your supply chain will collapse if you don't maintain the infrastructure." Shin looked at her. Direct, as always. "You're the only person who can do what you're doing here. If you burn out, everything stops."
"People keep telling me I'm important. Nobody tells me I have time."
"You have time. The crystal is growing for forty-eight hours. The Chinese contact hasn't been established yet. The third Mugyeong mission won't launch for at least a week. Tonight, you have time."
Sera leaned her head against the containment enclosure's outer shell. The metal was cool. Behind it, through two layers of mana shielding, the rat was sitting in its cage, pressed against the bars closest to her head. She could feel its presence — not through any scientific instrument, but through the awareness that came from sharing space with something you'd been studying for three weeks. The rat's mana emissions, usually undetectable without equipment, vibrated through the metal at a frequency her skin had learned to recognize.
"Tell me about Incheon," Sera said.
Shin was quiet for a moment. "The dungeon drop processing facility?"
"You worked there for two years. Hazardous materials handling. You processed monster parts — organs, fluids, bones, all the material that dungeon clearance teams brought back. What was it like?"
"Boring, mostly. Twelve-hour shifts, industrial processing, safety protocols for every step. We wore hazmat gear and processed materials on conveyor systems. Like working in a slaughterhouse, except the meat was from another dimension and could sometimes explode."
"Why'd you leave?"
Shin was quiet again. Longer this time. The kind of quiet that surrounded a memory being retrieved from a place it had been deliberately stored.
"An accident. A B-rank monster core came through processing with an incomplete deactivation. Standard procedure is to verify deactivation at three checkpoints. The first two confirmed dead. The third — my checkpoint — I flagged it as active. My supervisor overruled the flag. Said the instruments were miscalibrated. Pushed the core through to final processing."
"What happened?"
"The core reactivated during dismantling. The processing technician lost his right hand. The facility was shut down for two weeks. My supervisor was transferred. I was commended for the correct initial assessment and then reassigned because my presence made the new supervisor uncomfortable."
"Because you'd been right and your predecessor had been wrong."
"Because institutions don't reward the people who catch mistakes. They reward the people who prevent the embarrassment of mistakes being caught." Shin's voice was even. Not bitter — processed. A feeling that had been examined, understood, and filed. "I volunteered for Colonel Hwang's assignment because it was the first time someone offered me a position where being right wasn't a liability."
"And here you are. In a basement lab with an alchemist who blows things up and a rat that's evolving into something alien."
"Here I am." The ghost of a smile. "At least your mistakes are interesting."
They sat in silence for a while. The lab hummed. The crystal growth furnace ticked softly as it maintained temperature. Min-su stood in his corner, eyes closed but not sleeping — his breathing was too controlled, too regular. A man resting while remaining alert.
The rat shifted behind Sera's head. Its crystalline structures scraped against the cage bars with a soft chiming sound — like wind chimes, but lower, resonant, a tone that vibrated in Sera's chest.
"It's singing," Shin said.
Sera listened. The rat was dragging its crystalline foreleg along the bars in a deliberate pattern — not random scraping, but a sequence of tones. High, low, middle, low, high. A five-note pattern that repeated, each repetition slightly different, as if the rat was refining the melody.
"The crystals produce different tones at different contact angles," Sera said. "It's figured out that the cage bars create sound when the crystals touch them, and it's... experimenting. Learning which angles produce which tones."
"It's making music."
"It's exploring acoustic properties of its own biology interacting with its environment."
"Sera. It's making music."
The five-note pattern played again. Higher this time. Clearer. The rat had found the angles it wanted and was producing them with increasing precision. The tones drifted through the lab — small, delicate, alien sounds that had no business being beautiful and were anyway.
Min-su's eyes opened. He listened. His expression didn't change, but he listened.
Sera pressed her back against the enclosure and let the sound move through the metal and into her spine. The rat's composition — if that's what it was — continued for another three minutes. Five notes, arranged and rearranged, a crystal instrument played by a creature that had been a C-rank dungeon vermin twenty-three days ago and was now something that explored beauty for its own sake.
When it stopped, the silence was different. Heavier. Full of a sound that was no longer there but hadn't been forgotten.
"Day forty-one," Sera said quietly. "Twenty-two months minus forty-one days. And a rat in a cage just taught itself to sing."
She didn't write it on the tablet. Some things didn't belong in encrypted files reviewed by military intelligence officers.
Some things were just for the people in the room.
---
Kang called at 0700 on day forty-two with preliminary results from the crystal growth.
"The lattice is forming correctly. Doping is uniform. I'm seeing resonance onset at approximately 3.4 terahertz — close to target but not exact. The crystal needs another twelve hours of growth to reach full maturation."
"3.4 versus the target 3.7. Can you adjust the doping concentration to push it higher?"
"Not without restarting the growth from scratch. The doping compound is already at maximum concentration for a stable lattice. Pushing it further risks structural defects — the crystal would resonate at the right frequency but shatter when subjected to mana load."
"A crystal that shatters when you use it. Useful."
"The 0.3 terahertz gap may not matter for a proof of concept. The crystal will resonate in the divine-class frequency range, even if it's not precisely tuned. Whether that's sufficient for your purposes depends on how frequency-specific the recipe requires its components to be."
Sera checked [Brew]. The probability tree for the ability-code potion was still in her memory — mapped during the window, committed to organic storage that no System protocol could monitor. The divine-resonance crystal node had a frequency tolerance range.
"The recipe specifies 3.5 to 3.9 terahertz," she said. "3.4 is below the threshold."
Silence on the line.
"How far below?" Kang asked.
"Two percent. The crystal will be two percent below the minimum resonance frequency required for the recipe to function."
"And two percent is—"
"The difference between a key that fits the lock and a key that scratches the tumbler. Close enough to feel right. Too far to turn."
More silence. Sera could hear Kang breathing — the slow, measured respiration of a man calculating options.
"I can try a second growth," he said. "Different doping approach. Instead of the raw fluid, I use the tertiary compound — it has a higher resonance frequency because of the hybrid molecular architecture. If I dope with the compound instead of the fluid, the crystal might reach 3.7."
"The compound is Protocol-Restricted. Using it as a doping agent for crystal growth will trigger the System's new monitoring protocol."
"The System already knows you have the compound. Using it isn't a violation — it's an approved experimental material. The monitoring protocol tracks creation of new compound, not usage of existing stock."
He was right. The System's new protocol targeted the *methodology* — combining fluid with hybrid blood to create the compound. Using compound she'd already synthesized was different. The System would monitor the usage, would see the compound being consumed in a crystal growth process, but it wouldn't trigger the creation alert.
"Do it," Sera said. "Use the compound. How much do you need?"
"Two milliliters. Minimum."
She had six vessels. She could spare two milliliters. But every milliliter of the compound consumed was a milliliter she couldn't recover — the creation methodology was now monitored, and producing more would alert the System.
Finite resources. Permanent costs. The math of alchemy.
"I'll have it ready when you arrive," she said.
She hung up and looked at the six vessels of tertiary compound on the workbench. Dark. Violet. Irreplaceable.
Two for the crystal. The remaining four for the intermediate compounds and the final brew. Assuming the crystal worked. Assuming Kang could push the resonance above 3.5 terahertz. Assuming the proof of concept succeeded. Assuming the third Mugyeong mission produced the real crystal she needed for the full hack.
A lot of assumptions. Sera hated assumptions. She preferred variables — quantified, measured, controlled. Assumptions were variables without data, and variables without data were just hopes wearing lab coats.
She sealed two milliliters of compound into a transfer vessel, labeled it, and set it aside for Kang.
Then she went back to the combat potions, because the mission wouldn't wait for the crystal to grow, and soldiers needed tools that worked today, not breakthroughs that might work tomorrow.
