The Clean-Air Smuggler
Download The Clean-Air Smuggler as a PDF
The Dread Sanitarium · LV53 · Michael S. Moniz
Story catalogue · Reading guide
• • •
A clean-air cartridge is a cylinder the length of my forearm, sealed at both ends with hardite caps, containing approximately four hours of purified atmosphere compressed to a density that the lower hive cannot produce and the upper hive does not sell.
I carry six per run. Three in the pack across my shoulders, two in the belt holsters, one inside my jacket against the ribs, where the body heat keeps the compression stable during the descent. Six cartridges. Twenty-four hours of breathable time. The weight is manageable if you have learned how to distribute it across the stride so that the load does not shift when you change direction in the service corridors, where the turns come fast and the walls are narrow and a shifting load can cost you the half-second that determines whether you reach the next junction before the lattice sweep returns.
My name is Pell. I am Adhi. I was a carrier on the lower processing levels until three years ago, when a woman I knew asked me whether I could run. Not the word the Sheru use for running — exercise, recreation, the movement of a body through space for its own sake. The Adhi word for running means to carry something through a route that does not want you on it. The route is the obstacle. The running is what you do to the route.
I told her I could run. I did not yet know what I would be carrying.
• • •
The cartridges come from a Sheru workshop on the boundary between the Sheru quarter and the Ennesh market district. I do not know the maker’s name. I know the workshop by its corridor number and the timing interval of its service door — six seconds, wide enough for a runner with a full pack to enter and exit during the lattice reset. The maker is Sheru. The compression technique is Sheru craft — Tier One substrate work adapted to atmospheric containment. The same material that the ornament guilds use for decorative panels can be shaped into a pressure vessel that holds clean air at a density the lower hive has never breathed.
Clean air is not illegal. This is important. The air at the Sheru level is not a controlled substance. The air at the Ennesh level is not a controlled substance. Even the bright air of the Aenu platforms is not, technically, prohibited to anyone who can reach it. What is illegal is the distribution of atmospheric product outside the sanctioned ventilation infrastructure without a tariff licence. The cartridges bypass the system’s atmospheric allocation — the vertical distribution model that determines, through altitude and filtration, who breathes what quality of air. Moving clean air downward without a tariff is theft from the allocation model. The crime is not the air. The crime is the route.
I buy the cartridges from the maker at a price that reflects the material cost of the substrate casing, the labour of compression, and a margin that I do not negotiate because the maker is providing something the system does not offer and the market price for things the system does not offer is whatever the maker says it is. I sell the cartridges in the lower hive at a price that reflects the delivery risk, the route complexity, and the fact that four hours of clean air is four hours during which a processing-line worker’s lungs are not accumulating the residue deposits that will, over twenty years, calcify the tissue and reduce their breathing to the shallow careful inhalations of someone who has learned to ration something that should not need rationing.
I am selling time. Not metaphorically. The cartridges extend the functional lifespan of the lungs that breathe them. Four hours of clean air is four hours of recovery — four hours during which the tissue can process the accumulated waste without taking on more. The clients use the cartridges at night, when they sleep, when the body’s repair mechanisms are most active. They attach the cartridge to a simple mask — cloth and hardite, Adhi make, nothing the system would classify as medical equipment — and they breathe clean air while the lower hive breathes what it always breathes.
The mask is a strip of tightly woven fabric fitted over the nose and mouth, with a hardite coupling that connects to the cartridge’s output valve. The design is not mine. It circulated through the lower hive before I started running — a piece of practical engineering that someone made because someone needed it, passed from shelter unit to shelter unit the way useful knowledge moves through any community that the official system does not serve. The mask does not filter the ambient air. It bypasses it entirely. When the valve is open, the sleeper breathes only what the cartridge contains. For four hours, the lower hive does not exist. The heavy air, the residue, the taste of metal and spent geometry — gone. Replaced by the air that the Sheru quarter breathes without thinking about it. The air that the Ennesh markets consider unremarkable. The air that is, at the Adhi level, a luxury so extreme that the system has no tariff category for it because the system never imagined anyone would try to sell it.
In the morning they wake with a clarity in the chest that lasts until the shift begins and the heavy air fills them again. The clarity is not permanent. It is not curative. It is a reprieve — four hours during which the damage pauses and the body remembers what undamaged breathing feels like. The product is not health. The product is the memory of health. The memory is what they are paying for.
The product is four hours of being someone the system did not design you to be. The price is what the market will bear. The crime is that the market exists at all.
• • •
I run three routes. Each route descends from the Sheru-Ennesh boundary to the lower hive through a different set of service corridors — the maintenance passages, utility conduits, and structural access ways that connect the hive’s functional levels without passing through the classified thresholds that the geometric intelligence monitors.
The service corridors are not secret. They are infrastructure. Every hive city needs passages for maintenance crews, waste handlers, substrate delivery, and emergency access. These corridors are built into the hive’s architecture the way veins are built into a body — essential, ubiquitous, and largely ignored by the systems that monitor the arterial traffic. The lattice reads the main corridors with continuous attention. The service corridors receive periodic sweeps — lattice pulses that pass through the infrastructure on a timed rotation, reading whatever is inside at the moment of the sweep. Between sweeps, the corridors are unmonitored.
The corridors themselves are narrow. Not designed for comfort — designed for access. The walls are rough-cut stone without the geometric channelling that lights the residential and commercial levels. In the service corridors, you carry your own light or you know the route well enough to run it in the dark. I know my routes well enough to run them in the dark. I know where the ceiling drops at the junction between Sheru-3 and the upper maintenance spine. I know where the floor is wet from condensation that the decommissioned drainage system no longer removes. I know the seventeen steps down between the maintenance spine and the Adhi-level access shaft, and I know that the twelfth step is cracked and will shift under weight if you do not place your foot at the left edge where the stone is still bonded to the substrate beneath.
The corridors smell of old substrate and damp stone and the faint mineral residue that settles everywhere the ventilation does not reach. In the deepest sections — the connective passages between the Sheru quarter and the Adhi residential blocks — the air is appreciably heavier than the Sheru level above. You can feel the altitude change in your lungs before you feel it in your legs. The air thickens. The residue concentration increases. The taste of metal appears on the tongue. By the time I reach the lower hive, I am breathing the same air my clients breathe every day, and the cartridges in my pack are the only clean air within a hundred metres.
Timing is the skill. Not speed — the corridors are narrow and uneven and running flat-out through them would produce the kind of noise and air disturbance that the lattice’s secondary sensors detect even between primary sweeps. The skill is knowing the interval between sweeps for each section of each corridor, and moving through each section during the gap. The intervals vary. Some corridors sweep every ninety seconds. Some sweep every four minutes. The variance depends on the corridor’s classification — higher-priority maintenance access sweeps more frequently because the system assigns greater operational value to the infrastructure it connects. A corridor that services the Ennesh market district’s atmospheric systems sweeps more often than a corridor that connects two decommissioned storage blocks in the lower hive.
I learned the intervals the way the Adhi learn everything about the city — through the body, through repetition, through the cost of getting it wrong. The first time I mistimed a sweep I was carrying two cartridges in a belt holster through a utility conduit between Levels Sheru-3 and Adhi-7. The lattice pulse caught me in the open. It read my Adhi signature in a Sheru-access maintenance corridor carrying two objects whose geometric composition did not match any registered cargo classification. The system flagged the reading. I ran. By the time a maintenance inspection crew reached the corridor, I was four levels below and the cartridges were cached in a wall cavity I had identified on a previous run.
The flag stayed in the system. I do not know what the system did with it. Flags in the maintenance corridor data are not unusual — rats, structural debris, pressure fluctuations, and the occasional lost Sheru worker produce readings that the system flags and then deprioritises when no follow-up confirms the anomaly. My flag may have been deprioritised. It may not. I changed routes after that. I learned the intervals more carefully. I have not been caught in a sweep since.
But being caught in a sweep is not the real problem. The real problem is the air itself.
• • •
I will describe a delivery so that you understand what the work is.
I collect the cartridges from the Sheru workshop at the start of third shift, when the corridor traffic is lightest and the lattice sweeps in the service corridors are at their widest intervals — the GI reduces sweep frequency during low-traffic periods as an efficiency measure, which means the system’s most vulnerable window is also its most efficient. I load the pack: three cylinders, padded with cloth to prevent the hardite caps from striking each other during movement. I slot two into the belt holsters, where they sit against my hips and add a lateral weight that I compensate for by angling my stride slightly inward. I place the sixth inside my jacket, against the left side of my ribcage, where my body heat stabilises the compression.
The first descent is the easiest. From the Sheru-Ennesh boundary to the upper maintenance spine: a wide corridor with good footing, a sweep interval of three minutes and forty seconds, and enough ambient geomagic in the walls to provide faint illumination. I cover this section at a brisk walk. Running here would be wasteful — the interval is generous and the corridor is long enough that I would arrive at the junction with time to spare and breath to waste. Breath matters. Every exertion in the service corridors costs lung capacity that I will need later, when the air is heavier and the intervals are tighter.
At the junction I wait. The sweep for the lower maintenance spine runs on a different cycle — two minutes and ten seconds. I count. I have learned to count at a rate that matches my resting heartbeat, which gives me a consistent internal clock that does not drift with anxiety or exertion. Eighty beats. Ninety. At one hundred and twelve beats I enter the lower spine and move.
This section is darker, narrower, and descends at a grade that makes the pack shift against my shoulders. I adjust the straps as I walk. The air is changing already — the Sheru-level atmosphere giving way to the transitional zone where the ventilation systems from above meet the unfiltered air from below. The taste of metal appears. Not strong — not the heavy mineral density of the Adhi levels. A preliminary flavour. A warning that the body reads before the mind names it.
The seventeen steps between the maintenance spine and the Adhi access shaft. I take them quickly, placing my feet from memory, avoiding the cracked twelfth step. The shaft is vertical — a ladder built into the stone wall, descending four levels through the structural interior of the hive. The ladder is maintenance infrastructure, used by vent workers and waste handlers during service shifts. At third shift, it is empty. The lattice does not sweep the shaft continuously — it pulses at the top and bottom, with a gap of approximately forty-five seconds between pulses. Forty-five seconds to descend four levels on a stone ladder with six cartridges and a pack that wants to pull me backward.
I climb down quickly, hands and feet finding the rungs in the dark, the cartridges in the belt holsters pressing against the ladder with each step. The one against my ribs shifts. I press it with my elbow. The shaft smells of damp stone and old residue and the particular stillness of a space that no one enters voluntarily. At the bottom I step off the ladder into the Adhi-level access corridor and I am in the lower hive.
The air is immediately, measurably worse. The heaviness is not gradual at this transition — the shaft drops through the atmospheric boundary between Sheru-equivalent air and Adhi-baseline in the space of four levels, and the body registers the change as a weight in the chest, a texture on the tongue, a slight resistance in the breathing that will become normal within a few minutes and will remain normal for as long as I am down here. I breathe it. This is my air. I was born in it. My lungs are adapted to it. The adaptation does not make it comfortable. It makes it survivable.
The dead drop is two corridors from the shaft exit. I reach it in under a minute, deposit the cartridges in the wall cavity behind a loose stone facing, collect the payment that was left on the previous cycle, and turn back. The return climb is harder — the pack is lighter but the lungs are heavier, and the forty-five-second window on the shaft feels shorter when you are climbing against gravity with residue in your chest.
Total elapsed time from the Sheru workshop to the dead drop and back: twenty-two minutes. Six cartridges delivered. Twenty-four hours of breathable time placed into the informal economy of the lower hive. The system’s atmospheric monitoring has not been triggered. The lattice sweeps have not caught me. The delivery is invisible.
Except for the air. The air will remember.
• • •
When a clean-air cartridge is opened in the lower hive, the purified atmosphere mixes with the ambient air. The mixing is local — a radius of perhaps twenty metres from the point of use, depending on the ventilation conditions in the residential block. Within that radius, the atmospheric contamination drops. The residue density decreases. The air becomes measurably cleaner for the duration of the cartridge’s output — four hours, gradually diminishing as the compressed atmosphere depletes.
The geometric intelligence monitors atmospheric density at every level of the hive. This is not targeted surveillance. It is infrastructure management. The GI tracks contamination levels because contamination levels determine ventilation scheduling, filter maintenance, and the health assessments that confirm each caste’s environmental baseline. The monitoring is continuous, passive, and comprehensive. The GI knows what the air tastes like at every altitude because knowing the air is part of knowing the city.
A pocket of clean air in the lower hive is an anomaly. Not a dramatic anomaly — not the kind of pressure differential that a vent worker would flag as an urgent maintenance concern. A subtle one. A slight decrease in the ambient residue density within a residential block, lasting approximately four hours, concentrated in a radius consistent with a point source rather than a systemic change. One pocket, one time, falls within the noise margin of natural atmospheric variation. The GI notes it and moves on.
Two pockets in the same block within the same week is a pattern. Three is a data set. The GI does not investigate patterns. The GI records patterns and makes them available to the systems that do investigate — the maintenance scheduling algorithms, the health assessment protocols, the Uthra engineering division’s infrastructure auditing tools. If anyone with access to those tools asks the right question — why is the contamination level in residential block Adhi-9 consistently lower than the environmental model predicts on these specific nights? — the data is there. The data has always been there. My deliveries are invisible only as long as no one looks at the atmospheric record and sees the clean fingerprints I have been leaving in the dirty air.
Every successful delivery adds another data point. Every cartridge opened in the lower hive produces another four-hour pocket of anomalously clean air that the GI records with the indifferent thoroughness of a system that records everything and understands nothing. The data accumulates. The pattern sharpens. The route from the Sheru-Ennesh boundary to the point of sale in the lower hive is not written on any map, but it is written in the atmospheric record — a trail of clean air pockets descending through the hive like a line of stones across a river, each one showing where I have been and where my clients sleep.
I am selling breathable time. The cost of selling it is that every sale makes the next sale more visible.
• • •
My clients do not know my routes. They do not know my schedule. They know a contact in their residential block who knows a contact in the adjacent block who knows how to leave a marker at a dead drop in a service corridor that I check on my way down. The marker indicates demand — how many cartridges, which block, which shift cycle has the fewest people awake to notice the delivery. I leave the cartridges at the drop and collect the payment on the return run. I do not meet my clients. They do not meet me. The distance is not operational paranoia. It is practical recognition that the fewer people who can connect the cartridges to a specific runner, the fewer people the system can audit if the atmospheric data produces an investigation.
I know some things about them anyway. I know from the contact network that most of my clients are processing-line workers with ten or more years on the residue compression channel. I know that the cartridges are shared — a single four-hour cylinder passed between two or three people sleeping in the same shelter unit, each person taking an hour or two of clean breathing before passing the mask. I know that the children get the longest turns. I know that the demand increases after the health station’s quarterly baseline assessments, when the workers are told again that their lung function is consistent with their environmental classification and that no treatment is indicated. The workers already know this. They come to me because the system’s answer — your damage is normal — is not an answer they can live inside any longer. Not when four hours of clean air shows them what living outside that answer feels like.
I know that a woman in Block Adhi-9 has been buying two cartridges per month for eighteen months. I know this because her order is the most consistent in the contact network — never more, never less, always on the same rotation, always collected from the same dead drop. The consistency tells me she has budgeted for this. She has carved the cartridge cost out of a shift wage that barely covers her ration entitlement and her shelter contribution, and she has done this for eighteen months because whatever the clean air gives her is worth more than whatever the money would have bought inside the system’s approved economy. I do not know what the clean air gives her. Maybe it is the breathing. Maybe it is the hours of sleep without the cough. Maybe it is the knowledge that for four hours every other week, the system’s answer is wrong.
I also know that the demand is growing faster than I can supply. Three years ago I delivered twelve cartridges per month. Last month I delivered forty-one. The contact network has expanded to four residential blocks in the lower hive. The price has not dropped. The clients pay it because the alternative is free — the system provides all the heavy air you can breathe at no cost — and the alternative is killing them.
I cannot expand supply because expanding supply means more runs, more cartridges, more clean-air pockets in the atmospheric data, and a trail that becomes less a line of stones across a river and more a road that the geometric intelligence can read without being asked. The business model has a ceiling. The ceiling is the system’s capacity to notice what I am doing, and that capacity increases with every cartridge I sell.
• • •
Last week the contact in Block Adhi-12 told me that a vent worker on the lower-mid route had filed an observation about the atmospheric readings in the block. Not an anomaly report. An observation. A notation in the maintenance registry that the contamination levels on certain nights were lower than the environmental model predicted.
I thought about Vedra. Not a person I know — a type. Every lower-mid route has a vent worker who notices when the numbers stop reconciling. They notice because noticing is what they are trained to do. The filters, the pressure differentials, the resonance readings — the vent worker’s entire profession is the detection of atmospheric irregularities. My deliveries are atmospheric irregularities. I have been creating the exact kind of data signature that a competent vent worker would identify as an anomaly — a pattern in the readings that does not match the system’s model. The vent worker is not my enemy. The vent worker is doing their job. Their job and my job cannot coexist in the same atmospheric data set. I produce clean-air pockets. They detect clean-air pockets. We are opposite functions of the same system.
The observation is in the registry now. It has not yet been escalated. It may not be. Observations sit in the maintenance registry alongside thousands of other notations about pressure fluctuations, temperature drift, and the small inconsistencies that the lower hive’s aging infrastructure produces daily. My clean-air pockets may disappear into the noise. They may not. The difference depends on whether the vent worker files a follow-up, whether the follow-up coincides with another data point, whether the geometric intelligence’s pattern-matching protocols are triggered by the accumulation of observations in the same residential area.
I cannot control any of these variables. I can control how many more data points I add.
I have stopped deliveries to Block Adhi-12. I have reduced the schedule across all four blocks — from forty-one cartridges per month to twenty-two, delivered on a randomised rotation designed to break the temporal pattern in the atmospheric data. I have changed two of my three routes. The new routes are longer, slower, and pass through corridors whose sweep intervals I have not yet fully mapped. The risk per delivery has increased. The total deliveries have decreased. The clients in Block Adhi-12 are breathing the air the system provides. The heavy air. The air that is consistent with their environmental classification.
A man in Block Adhi-12 sent a message through the contact network asking when the deliveries would resume. He said his daughter’s cough had returned. He said it had been gone for four months — four months of sleeping with the mask, four months of clean air rebuilding what the heavy air had damaged. He said the cough was worse now than before the cartridges, because now her lungs knew what they were missing.
I do not know his name. I do not know his daughter’s name. I know that she is breathing the air the system designed for her, and that I gave her four months of something better, and that the four months made the return to the baseline worse. I gave her lungs a memory of clean air. The memory is the cruelty. The lungs that have been repaired — even partially, even temporarily — do not return to their previous tolerance. They have tasted what they were meant to breathe. The heavy air is the same air it always was. The lungs are not the same lungs.
The woman in Block Adhi-9 — the consistent client, the one who has budgeted for two cartridges per month for eighteen months — she is still receiving deliveries. Her block has not been compromised. Her atmospheric data, spread across a wider area and a more randomised schedule, has not yet produced the density of anomalies that would trigger a maintenance observation. She is still breathing clean air every other week. She does not know that Block Adhi-12 has been cut off. She does not know that the reason she still receives her cartridges is that another family has lost theirs.
This is the economics of the route. Every cartridge I deliver to Block Adhi-9 is a cartridge I am not delivering to Block Adhi-12. Every clean-air pocket I produce in one location is atmospheric data I am not producing in another. The supply is finite. The demand is not. And the system’s capacity to detect the supply grows with every unit I sell. I am managing a shrinking corridor between what I can deliver and what the system can see, and the corridor narrows every month.
• • •
The service corridors are still there. The sweep intervals have not changed. The maker on the Sheru-Ennesh boundary is still producing cartridges — the compression technique is simple, the substrate is available, the demand has not decreased even as the supply has contracted. Twenty-two cartridges per month still descend through the hive’s infrastructure, still produce four-hour pockets of clean air in residential blocks that the system monitors with the passive thoroughness of a machine that was built to know what the air tastes like everywhere, all the time, forever.
I am still running. I will always be running. Not because I am brave or ideological or committed to a cause that transcends my own survival — I am none of these things. I am Adhi. I am a carrier who learned to run. I run because the demand exists and because the product is air and because the system’s answer to the people who breathe at my altitude is that their breathing is consistent with their classification and that no intervention is indicated. I run because the alternative to running is agreeing with that answer.
But I am running slower. Twenty-two cartridges instead of forty-one. Randomised rotations instead of efficient schedules. New routes with unmapped intervals instead of the corridors I knew by heartbeat. Every delivery I have ever made is written in the atmospheric record — a history of clean-air pockets that will remain in the GI’s data for as long as the system stores data, which is forever. The trail does not fade. The data does not expire. The system will have every data point I have ever produced, waiting for someone to ask the question that makes the trail visible.
I am running into a corridor that is narrowing. The system is not chasing me. The system does not need to chase me. The system is patient and comprehensive and it records everything and eventually the accumulation of data points — clean-air pockets, atmospheric anomalies, maintenance observations, pattern flags — will reach the density at which the trail becomes a route and the route becomes an investigation and the investigation becomes a classification. Runner. Smuggler. Tariff evader. Atmospheric fraud. The system will find a category. The system always finds a category.
Until then I run. Twenty-two cartridges per month. Three routes. Six cartridges per delivery, the weight distributed across the stride, the pack across the shoulders, two in the belt holsters, one inside the jacket against the ribs where the body heat keeps the compression stable. The woman in Block Adhi-9 will receive her two cartridges this cycle. The man in Block Adhi-12 will not. His daughter will cough. The system will classify the cough as consistent with her environmental baseline. The system will be correct.
The service corridor is dark. The sweep interval on this section is one hundred and fourteen seconds. I have eighty-seven seconds remaining before the lattice pulse returns. The air down here is heavy. The air in my pack is clean. The distance between the two is a service corridor, a timing window, and a crime that consists entirely of giving someone permission to breathe.
I run.
• • •
Fiction by Michael S. Moniz. Produced with AI assistance in writing and editing under the author’s direction. H.P. Lovecraft is the literary influence; references to Lovecraft in source credits identify the inheritance or AI writing persona.
Edition and source
Selected final: LV53_THE_CLEAN_AIR_SMUGGLER.docx. Selection basis: only version. Story prose preserved; reading-edition title and credits standardized, source word-count notes omitted.