I began mapping in earnest on the tenth day.
Not the casual geography I had been building through movement — the rough mental model of rock formations, canyon positions, resource zone locations. Actual systematic mapping. I walked circuits around the region I had been operating in, expanding the radius each time, noting every observation point, every convergence zone I could detect through the active absorption modification, every population concentration.
The Infernal Realm, at least this section of it, was not random.
That was the first thing the mapping revealed. The rock formations — the enormous spires, the canyon cuts, the flat elevated shelves between them — had a pattern that was not geological in any process I recognized from Earth. The distribution of resource zones was similarly non-random: roughly equally spaced, with the primary canyon I had been working representing the highest-density zone within the radius I had covered. The spacing suggested either deliberate design or a self-organizing system that had settled into this arrangement over very long periods.
The demon population distribution followed the resource zones. The weakest, most numerous demons occupied the spaces between zones. The stronger demons occupied the zones themselves, in numbers that their resource density could support without mutual destruction — which suggested the zones maintained rough equilibrium over time, predator-prey dynamics operating at a power system level.
I had been treating this environment as chaos. It was not chaos. It had structure.
The implications were positive. Structure was learnable. Structure had predictable zones of higher and lower risk. Structure meant I could plan movements rather than react to surprises.
The negative implication was that I was not the only entity capable of reading structure. Whatever occupied the upper tiers of this world's hierarchy had either built this structure or was maintaining it, and I was operating inside a system that had been designed by or for something far above my current capability.
Something to remember.
I found the anomaly on the thirteenth day, in the southern wall of the primary canyon.
I had been making careful sweeps of the canyon's perimeter — never entering, maintaining observation distance, tracking the advanced entity's movements and building a predictive model of its patrol range. What I had established so far: the entity moved in a roughly circular pattern around the primary convergence point, with a radius that varied between fifty and eighty meters. It spent the longest periods at the convergence point itself, absorbing. Its circuit took approximately three hours. It was regular enough that I could be on the northern rim when it was at the southern end without being in its immediate attention range.
The southern wall, during one of these observation windows, showed me something that had no place in the geology.
It was not visible in any obvious sense. There was no glow, no distortion of the air that you could see directly. What I perceived it through was the energy — the same way I had learned to feel the convergence gradients, the same ambient sensitivity that the active absorption modification had enhanced. At one section of the southern wall, roughly forty meters above the canyon floor and twelve meters wide, the energy behaved differently.
Everywhere else, the ambient demonic energy flowed downward, toward the convergence sources below the surface. At this section of the wall, the energy moved laterally — passing through the rock as if the rock were not there, flowing from somewhere beyond it in a direction that had no correspondence to anything in the local environment.
Something was on the other side of that wall.
Not another section of the Infernal Realm. I had enough sensitivity now to read the character of local demonic energy, and what was flowing through the wall's anomaly had a different character entirely. Not wrong, exactly. Just different — the way water from two separate sources, mixed, still tastes of both.
I spent four hours watching it from the canyon rim.
The flow was not constant. It pulsed — slow, irregular, without the rhythmic regularity of a biological pulse. More like breathing during sleep: variable, responsive to something internal, pausing and resuming without apparent external trigger. During the stronger pulses, I could feel the foreign energy touching my own reservoir through the absorption modification and being drawn in passively — and when it was, it felt distinct. Cleaner, in some way. Higher-frequency, if that description applied to energy.
A portal.
Unstable. Not open — there was no visible aperture, no crossing point I could have physically stepped through even if I wanted to. But the mechanism was there, in a partial or dormant state. The wall between this section of the Infernal Realm and whatever was on the other side was thinner here than it was anywhere else I had found.
I did not approach it. Not because I was afraid, but because I had a rule that I had been developing since the first night in the dark with the system: never interact with something you don't understand before you understand it. Observation cost nothing. Premature interaction could cost everything.
I marked the location in my mental map with as much precision as I could achieve, noted the pulse pattern with its approximate timing intervals, and filed it under long-term investigation with a note: access to other worlds through this mechanism — study first, use when ready.
The system had told me the Infernal Realm connected to infinite worlds via unstable portals. This was the first one I had found. It was not the last I expected to find, and it was not yet the resource I would eventually use it as.
But knowing it was there changed the shape of my planning.
The hunting pack found me on the fifteenth day.
I want to be precise about that verb: found. They were not hunting me specifically — I was not significant enough in the local hierarchy to be a target of deliberate pursuit. What happened was that I crossed into their operational territory at the wrong time, and their behavioral pattern included any lone entity smaller than their dominant member as a valid resource.
I had been moving south of my established range, testing the edge of the primary canyon's energy gradient — where it faded enough that the advanced entity wouldn't read my absorption as competition — when I heard the movement pattern that was wrong.
Not one entity moving. Multiple. And the coordination implied by the sound pattern was not the loose proximity of entities that happened to be near each other. It was structured: two separate sound signatures flanking, a third heavier one behind, all moving in my direction at a pace that was deliberate.
I stopped.
Turned in a slow circle, using both hearing and energy sensitivity. The energy signatures were useful here: I could detect demonic energy as a presence now, and three presences, distributed around my position at roughly 120-degree intervals, was not coincidence.
I climbed.
The rock formations in this area were lower than the spires near the canyons — more like ridges, three to five meters high, with traversable surfaces if you moved carefully and with a body that had the bone density and claw structure I had been developing. I reached the top of the nearest ridge in approximately eight seconds and turned to look.
Six of them.
Not three. The flanking pair and the heavy rear were the most obvious. The other three had been lower to the ground — moving in the gaps between rock formations, below the acoustic profile that would have been easy to detect.
Coordination. Real coordination, not instinct.
I had miscounted the pack because someone or something had designed the approach to be miscounted.
Six strong demons, moving in a formation that indicated group hunting experience, surrounding a position at roughly twenty-five meters from my current elevated location.
I ran the assessment in the time it took them to register that I had moved vertically.
Six strong demons, 100 souls each. Six hundred souls in this group. Against my 263 current reservoir, the surface layer modification, the absorption modification, and the focused projection. My body: enhanced bones in the forearms, hardened nails, extended and weighted tail, baseline graphite skin with intermediate keratin-chitin composition.
What I had going for me: elevation, foreknowledge of the pack size, charged reservoir, and no obligation to stand and fight every one of them simultaneously.
What I had against me: six opponents with apparent coordination experience, unknown individual capability variation within the strong demon tier — I had only fought solo strong demons and could not assume these had identical parameters — and the fact that six simultaneous attackers changed the combat mathematics completely.
One-on-one, I could manage a strong demon by analysis, positioning, and the projection opening. Six-on-one, the projection would be spent in seconds, and I would be in melee range of multiple attackers before I could reload enough energy for a second round.
They were below me and had not yet determined which direction I had moved. I had perhaps ten seconds before the elevated position was obvious.
I made a decision.
Not all six. Two.
I selected the two closest — a pair that had been the left flank, currently positioned almost directly below my ridge, closer to each other than to any other member of the pack. I selected them because proximity meant I could engage them before the others closed distance, and because tight proximity to each other was a mistake I could exploit.
I jumped down behind them, not in front.
The drop was four meters. This body handled it without thought, the tail adjusting balance automatically on landing, knees absorbing more than they should have been able to, and I was already moving before full contact registered. Surface layer: active. Reservoir: full from the last absorption session. Projection: charged.
The first demon turned toward me, registered me at two meters, and I put a projection into its face at point-blank range.
Point-blank changed the dynamics significantly. At twelve meters, the projection had excavated five centimeters of rock. At two meters, with the same energy output, the force had not dispersed at all. The demon's head was not rock. The result was immediate and decisive, and the demon was gone as a threat in the same moment I was already turning to the second.
The second was smarter than the first — or its reaction time was better — because it had already begun its attack before I finished turning. The strike caught me on the left shoulder, the same shoulder that had taken damage in my first fight, now protected by the forearm bone modification's edge effects. Not full protection. The impact registered through the surface layer and the underlying structure as a compression I felt all the way to the collarbone.
I did not fall.
I used the momentum of the hit to rotate into the demon's space rather than away from it — the tail providing the counterbalancing pull that made the rotation possible without losing my feet. Inside its reach. Too close for it to wind up again. I drove the hardened nails of my right hand into the softer tissue at its throat in three rapid strikes and shunted the minimum viable projection output through my palm at contact range.
Contact-range projection was not designed use. I had not designed it for that. What happened was therefore in the category of interesting unintended behavior rather than planned capability: the energy, unable to travel the twelve-plus meters it was optimized for, discharged directly into the biological tissue in contact with my palm, and the demonic energy interaction I had first noticed during a grip four days ago happened at projection scale.
The demon did not die cleanly.
I had four seconds before the remaining four members of the pack were on me, and I spent two of them confirming the second demon was not recovering and the other two moving.
Not toward the remaining four.
Away.
I ran.
This is worth clarifying: I did not run because four strong demons was automatically beyond my capacity. I ran because four strong demons with demonstrated pack coordination, now with confirmed knowledge of my position and the death of two of their members, would not approach me individually. They would approach simultaneously, and simultaneously was the variable my current build could not handle. The projection was a single-target tool at this stage. The surface layer bought seconds, not minutes. The tail and bones and nails were melee assets that required positioning I could not maintain under four-directional pressure.
The decision to disengage was not retreat. It was not even caution in the conventional sense. It was accurate mathematics.
They followed.
Faster than I had expected. Pack hunting had optimized their pursuit behavior — they spread as they ran, attempting to re-establish the flanking pattern, which was sophisticated enough that I felt something I would reluctantly classify as impressed. Someone had trained these entities, or they had developed genuine tactical memory through experience.
The formation told me something else: this was not a feral pack. Feral pack animals pursue in a cluster. These were spreading deliberately to cut off my options. This had hierarchy. There was a coordinator.
I identified it while running — the largest of the remaining four, staying toward the center rear of the pursuit formation while the three others flanked. Directing, not chasing. If I killed the coordinator, the formation logic collapsed. The remaining three might continue the pursuit on instinct, but without the pattern-maintenance function, they would cluster rather than spread.
The coordinator was the hardest target. Positioned farthest back, protected by the three ahead of it. That was intentional.
I needed to create the geometry where the coordinator was temporarily separated from its protection.
I found what I needed in the rock formation ahead: a narrow gap between two ridges, perhaps three meters wide and twenty meters long, with walls too high to scale quickly. A fatal funnel for me, if I entered and they followed. Fatal funnel, inverted, if only one of them followed and the gap geometry worked in my favor.
I calculated the coordinator's distance and speed. The three flankers were ahead of it and would reach the gap first if I went through. The coordinator would be briefly alone at the gap's entrance for two to three seconds while its flankers navigated the interior.
Two seconds.
I went through the gap at a full run, hit the far end, turned completely, and waited.
The first flanker came through and I let it pass — it was not the target.
The second came through — not the target.
The third came through — not the target.
The coordinator appeared at the gap entrance.
Two seconds alone.
I sent four projections in sequence — not four individual aim-and-fire sequences, I did not have the time for that — but a rapid cascade, the way you can string keyboard inputs when muscle memory handles the mechanics and conscious attention only picks the targets. Four outputs in approximately 1.2 seconds, each aimed at the coordinator's torso and head at a gap distance of fifteen meters.
Three of the four connected.
The coordinator went down.
I did not wait to confirm it was dead. I turned and engaged the three flankers who had turned back at the sound of the impact.
What followed was the hardest six minutes of my existence in the Infernal Realm so far.
Three strong demons, no coordinator, but still moving with the residual instinct of a pack pattern even without direction — not clustering completely, maintaining loose triangulation by habit. My reservoir was down to perhaps 40% from the projections expended. The surface layer was draining it further at the passive rate. My left shoulder had a deep compression injury that was limiting the rotation range on that side.
I could not keep all three at distance. At some point, two of them would close to melee simultaneously, and I would have to manage two at once while the third reset.
I chose the smallest of the three and killed it first.
Not the weakest. In a pack with a coordinator, the smallest might be the fastest, the most precise, the utility member that filled gaps the larger ones created. I chose it because it was between me and the other two, which meant engaging it first created a brief geometry where I had my back to a ridge wall and only two attack vectors instead of three.
I took hits during that fight that I had not taken in any previous engagement. The shoulder injury compounded — each impact on the left side built on the compression already there, and by the third exchange with the smallest demon, the arm was functionally limited to 60% capacity, the rotation gone, the projection from the left palm unavailable due to the tremor the injury had introduced.
But I killed it. Clean. Neck. The nails were doing real damage now, the hardening modification meaning each raking strike removed tissue rather than sliding off the thick demon skin.
Two left.
Reservoir: 25%. Surface layer: still running, probably a mistake at this point. I killed the layer and shifted to absorption mode instead, pulling in ambient energy while I circled the two remaining demons. They watched me change tactics with what I was increasingly certain was genuine intelligence — they did not charge during the pause. They waited.
Waiting for what?
The answer arrived thirty seconds later: me to run.
They knew their coordinator was dead. They were not pursuing a pack objective anymore. They were standing and waiting for me to decide this wasn't worth the cost and leave. A preservation calculation. Two demons that had seen two of their pack members die in direct confrontation with something they couldn't classify were not going to charge without good reason.
That was extremely interesting information.
These entities could make preservation decisions under pressure.
It also gave me options I wouldn't have had if they had simply charged.
I absorbed for thirty seconds, pushing the reservoir to 40%, then released the absorption mode and shifted back to expression. I extended the surface layer to my hands, and I walked toward them.
Not running. Walking. Deliberate.
They broke at five meters.
They ran.
I let them go.
The calculation was simple: four dead, two fled, 400 souls available from the kills, my shoulder needing recovery time, and the energy I would spend chasing two fast demons through terrain I knew less well than they did was not worth the 200 additional souls. The 400 was the better outcome with current resources.
I collected.
[ SOUL COLLECTED ]
Type: Strong Demon — Approximate Value: 100 SoulsSoul Reservoir: 363 / Unlimited
[ SOUL COLLECTED ]
Type: Strong Demon — Approximate Value: 100 SoulsSoul Reservoir: 463 / Unlimited
[ SOUL COLLECTED ]
Type: Strong Demon — Approximate Value: 100 SoulsSoul Reservoir: 563 / Unlimited
[ SOUL COLLECTED ]
Type: Strong Demon (Pack Coordinator) — Approximate Value: 130 SoulsSoul Reservoir: 693 / Unlimited
The coordinator had been worth more. 130 souls against the standard 100. I noted that soul value was not strictly flat within tiers — something about the coordinator's role or development had elevated its individual worth.
693 souls.
I sat on the ground among four dead demons and a shoulder that was going to be a problem for the next several days and looked at that number for a long time.
The pack had been the most dangerous situation I had yet survived. That was honest. The two survivors running had been a choice on their part, not a victory I had forced. If they had charged instead, the fight would have had a different probability distribution and I was not going to pretend otherwise.
I needed to fix the tactical problem that had created the risk.
The problem was simple in statement: I could not handle multiple simultaneous attackers at current capacity. The projection was single-target. The surface layer bought time but not victory. The melee capability, enhanced as it was, did not scale to managing three or more opponents at once.
What solved multiple simultaneous attackers?
Area effect.
A projection that covered more than a single target. The system had told me during my energy design review that an omnidirectional release was wasteful and indiscriminate — my own analysis, during the design of the focused projection. But there existed a spectrum between a focused single-point beam and a completely omnidirectional explosion. An arc. A cone. A sweep.
I entered Design Mode.
Modification addition: Projection mode — arc expressionBased on: Existing palm-channel projection architectureChange: Instead of narrowed single-point conduit, widen exit aperture to 60-degree arcTrade-off: Reduced per-target force at same energy cost; wider coverage areaMechanism: I understand nozzle geometry and how changing aperture angle affects the distribution of output. The same energy leaving a wider aperture distributes its force across the coverage arc rather than concentrating it at a single point.
The system assessed:
[ DESIGN ASSESSMENT ]
Modification: Projection Mode — Arc Expression (60-degree)Comprehension: Intermediate (Fluid dynamics applied; mechanism understood as aperture variation on existing architecture)Quality: GoodCost: 40 SoulsNote: Arc expression will not kill at typical strong demon tier at standard energy levels — force distribution reduces per-target impact significantly. Optimal use: suppression, interruption, crowd control. Single-target projection remains available as separate mode.Note: Mode switching requires 0.5 seconds between arc and focused.
Good quality. My comprehension was improving, and the system was reflecting that honestly.
40 souls. Leaving 653.
I approved it.
The arc modification built on the existing architecture the way I had designed it to — not a new conduit, but a change to the exit aperture's control mechanism. When I tested it: both palms could now express a wide fan of energy, low-force-per-point but covering a sixty-degree spread in front of me. Testing it against three points on a rock face simultaneously produced clear impact marks at all three, with individually lower penetration than a focused strike but collectively present.
Crowd suppression. Not crowd killing. The distinction was critical and I had no illusions about it.
But suppression changed the dynamics of a pack encounter significantly. If I could force multiple attackers to react to an arc sweep, I could break formation, create openings, and handle the resulting chaos with focused projections rather than fighting a coherent coordinated attack.
It was not a solution. It was a tool.
Correct tools in the right sequence.
The shoulder needed time.
I spent three days at the secondary convergence zone, running passive and active absorption alternately, sitting in Design Mode and doing nothing but thinking. The shoulder was healing — faster than human baseline, consistent with the metabolic rate the system had identified — but the compression injury was deep enough that rushing it was a bad risk.
I thought about the pack.
Specifically, about the coordinator.
The coordinator had been a strong demon with 130 soul-value. Its distinguishing characteristic had been behavioral: it directed the others. It positioned itself protectively. It made tactical decisions under pressure. The pack had operated more effectively as a group than the sum of its parts would suggest — their combined effort had nearly cornered me, which six individually-acting strong demons almost certainly would not have achieved.
That was the value of coordination. Not power. Not individual strength. The ability to make individual power coherent.
I filed that under things I should eventually be able to do with creatures I design.
The system allowed creature creation. I had not yet explored it — the soul cost for designed creatures was likely substantial, and my comprehension of what made an effective creature was incomplete. But the pack had given me a data point: coordination multiplied effectiveness dramatically. A small number of well-designed coordinated entities could outperform a large number of independent entities of equivalent total power.
The system had told me: creation of creatures. Not creation of minions or creation of soldiers. The word was creatures — implying something with independent function. Something designed rather than evolved.
I began a low-commitment design sketch in the workspace — not for instantiation, just for comprehension building. What would a creature optimized for coordination and pack function look like? What were the essential components?
The exercise revealed gaps in my understanding almost immediately. I knew what I wanted the creature to do — communicate with others, hold formation, assess threats and relay that assessment — but I didn't know what biological architecture produced those functions. A brain capable of tactical real-time coordination was not something I understood at a mechanistic level. I could describe the behavior. I couldn't describe the organ that generated the behavior.
Future problem. Not today's problem.
Today's problem was the shoulder, and the language I still didn't have, and the 653 souls sitting in the reservoir that needed to be spent on the right things.
I made a list.
Priority 1: Increase energy reservoir capacity. The limiting factor in the pack fight had not been skill or modification — it had been energy. I had run to 25% during the critical phase and been forced to pause absorption to recover before the final engagement. A larger reservoir meant more sustained output. The system had noted energy density modification as requiring 400 souls minimum. I now had 653. I could approach that design.
Priority 2: Left arm structural repair. The compression injury would heal, but I wanted to understand the healing process well enough to determine whether the system could accelerate or improve it. More importantly, I wanted to modify the shoulder's structural resistance — the injury had revealed a vulnerability at the joint that the bone density modification hadn't fully addressed.
Priority 3: Language. The advanced entity had spoken to me and I had lost the information. I was operating blind in a dimension that everything above my current tier apparently used. The system could not shortcut zero comprehension — I needed exposure. Finding exposure required approaching entities that used language without dying in the process.
Priority 4: Understanding what was below.
I arranged the list in execution order and began working through it.
The energy density modification design was the most technically demanding thing I had attempted in the system so far.
The soul reservoir was not a physical structure — the system had never described it in physical terms, and my own perception of it was abstract rather than located. It was not somewhere in my body. It was more like a property of my body, the same way a battery's charge is a property of the battery rather than a thing inside it.
Increasing a property that wasn't a thing was conceptually different from extending a tail or hardening a fingernail.
I spent two days in Design Mode on this before I attempted anything.
What I worked out: the reservoir was defined by how much demonic energy my fundamental biological architecture could hold before the excess radiated out rather than being retained. It was a saturation limit, like how water has a specific heat — there is a maximum amount of thermal energy a given mass of water can hold at a given temperature before it changes state.
What increased saturation limits in other systems?
Changing the medium. Increasing the medium's density or bonding capacity. Adding something to the medium that increased its capacity to hold the energy.
I understood none of those mechanisms at a biological level. But I understood the concept, which the system had told me was the prerequisite.
I tried a design built on conceptual compression: increase the demonic energy saturation density of all body tissue by uniform enhancement of tissue-level energy bonding capacity.
The system's response was immediate:
[ DESIGN ASSESSMENT ]
Modification: Demonic Energy — Reservoir Density Enhancement (Uniform)Comprehension: Conceptual (Mechanism inferred by analogy; biological detail absent)Quality: Basic-IntermediateCost: 410 SoulsNote: Conceptual-level comprehension produces functional but inefficient result. Estimated capacity increase: 40% above current baseline. Biologically seamless integration not guaranteed. Host may experience periodic discomfort as tissue adapts.Note: This is the minimum viable version of reservoir expansion. Future designs based on deeper comprehension will be more efficient and produce greater capacity increases per soul spent.
410 souls. 40% increase. Basic-intermediate quality.
This was expensive for what it delivered, and the system was honest about why: my comprehension of the mechanism was still at the analogy level. I was paying a premium for conceptual-level understanding instead of mechanistic understanding.
But 40% more reservoir was still 40% more reservoir. The pack fight had shown me what happened when I ran low at a critical moment. 40% more buffer was a genuine tactical improvement.
I approved it.
693 minus 410: 283 souls remaining.
The modification was the worst installation experience I had yet had. Periodic discomfort had been the system's phrasing, and I had translated that in advance as some amount of pain that I would note and file. What it actually produced was a three-hour period where every cell in my body was doing something simultaneously — not painful in the sharp localized way of an injury, but in the deep, pervasive, slightly nauseating way of a systemic process operating at full intensity across every tissue simultaneously.
I sat at the convergence zone and absorbed and did not move during those three hours.
When it settled, I tested the new capacity: I drew the reservoir to full and felt the ceiling. It was higher. Approximately where the system had predicted — 40% above the previous level. The full reservoir now felt like significantly more potential than it had before, the energy denser and more pressurized in a way that I expected to translate into more sustained output before depletion.
Good.
283 souls remaining.
I designed the shoulder repair. This one was simpler — I had been building comprehension of the joint architecture from the inside during the healing process, paying the same kind of close structural attention to my own body that I had paid to the demonic energy system. The joint had a gap in its defensive geometry: the ball-and-socket structure was well-supported on the anterior and posterior faces but relatively exposed at the superior aspect — the top of the joint — which was where the impact had compressed.
Modification: Shoulder joint — superior capsule reinforcementStructure: Dense fibrous capsule extension at superior aspect, analogous to existing anterior capsule; additional tendon attachment optimization to distribute force across wider anchor pointsUnderstood: Joint architecture (observed from inside during recovery); fibrous capsule function (understood from general anatomy knowledge — it holds the joint together through tension); tendon force distribution (mechanical principle, well understood)
[ DESIGN ASSESSMENT ]
Modification: Shoulder Joint — Superior Capsule ReinforcementComprehension: Functional-Intermediate (Structural mechanics well understood; biological material properties inferred)Quality: GoodCost: 28 Souls
28 souls. Good quality. The system was rewarding three weeks of gradual comprehension improvement.
283 minus 28: 255 souls remaining.
The installation was clean and precise, the best modification experience I had had. The joint felt immediately more stable when I rotated the shoulder — not dramatically different in normal range, but with a solidity at the end of the rotation arc, at the extreme of reach, that had not been there before. The superior aspect was no longer the weak point.
I checked the full shoulder range. Complete. The compression injury had healed ahead of the modification schedule, and the reinforcement was now installed over healthy tissue rather than compromised.
255 souls.
Language remained.
Finding language exposure without dying required finding an entity above the strong demon tier that was not immediately hostile and was, for some reason, present in a location where I could observe it from a survivable distance over a long enough period.
The advanced entity in the primary canyon was my best candidate. It had chosen not to engage me. That was not the same as being safe to observe at close range, but it was better than any alternative I had identified.
I needed to get closer without triggering its threat assessment.
The solution I arrived at was energy masking.
The surface layer modification had a secondary function I had been exploring since the shoulder recovery: I could run the layer at a much lower output than its designed expression level, below the threshold of easy detection, as a kind of reduced-signature mode. Strong demons had not been sensitive enough to detect the layer at this setting during any of my canyon observations. The advanced entity was a different order of sensitivity.
But there was a significant difference between expressing energy that attracted attention and reducing my energy signature to blend with the ambient field. If the layer was at the right output level — matching the ambient energy density rather than exceeding it — it would not broadcast my presence. It would instead help me match the background.
Camouflage, in energy terms.
I could not mask the demonic energy in my body itself — that was structural, not something I could suppress. But I could make the boundary between my body's energy field and the ambient field less defined. Less distinct. Harder to read as a point-source of demonic energy rather than a feature of the ambient field.
I designed the addition carefully:
Modification: Energy signature reduction — ambient field matchingFunction: Active adjustment of surface layer output to match local ambient energy densityMechanism: The absorption modification already measures ambient density (necessary for active intake). Use that measurement to calibrate surface expression to matching level. Net effect: the energy boundary between my body and the environment becomes less defined.Note: This does not hide me. It makes me less distinct from the background. An entity with high enough sensitivity will still detect me. The goal is to reduce detection range, not achieve true invisibility.
[ DESIGN ASSESSMENT ]
Modification: Energy Signature — Ambient Matching LayerComprehension: Good (Mechanism well understood from existing modification architecture; sensing and expression already integrated)Quality: Very GoodCost: 32 SoulsNote: Effective against entities with standard or moderate energy sensitivity. Advanced entities with high sensitivity: partial reduction only. Against the highest-tier entities: minimal effect.
Very Good quality. The system reflecting that I was getting better at this.
32 souls. 255 minus 32: 223 remaining.
The modification integrated smoothly. When I activated it, I felt the surface layer shift from its standard broadcasting quality to something quieter — a change in the energy's relationship to the ambient field that I could feel from the inside as a kind of settling, as if I were exhaling.
I tested it against the only reference I had: the pack survivors. The two demons that had fled the fight four days ago. I had been tracking their location loosely through energy signatures — they had not left the region, which told me something about how pack territories worked. When I moved within thirty meters of one of them while running the ambient matching layer, it did not react until I was at approximately fifteen meters.
Without the layer: strong demons had noticed me at twenty-five to thirty meters during the pack hunt.
The reduction in effective detection range was real.
Against the advanced entity, the system had said: partial reduction only. I would plan for it to notice me at half its standard detection range rather than the full amount. Not invisible. Less visible.
Enough to get closer.
I approached the primary canyon on the eighteenth day with every modification active in their quietest modes: absorption running, surface layer at ambient-matching minimum, shoulder reinforcement settled, projection charged and ready.
223 souls in the reservoir and the expanded reservoir's 40% additional capacity above that.
I chose a position on the canyon's southern wall — not the rim, but a cleft in the wall face itself, approximately ten meters above the canyon floor and forty meters from the anomaly I had identified. The convergence lines were strong here. Stronger than I had felt at the secondary zone. The absorption rate at ambient-matching minimum was still pulling in more than I was expending to maintain the layer.
I was running a net positive energy state while hiding from an advanced entity.
I settled into the cleft and waited for the entity to complete one circuit. It came past my position at the end of its sweep — within twenty meters of me — and did not stop, did not look directly at the cleft, but its pace changed momentarily. A fractional hesitation. Then it continued.
It had detected something. Not me specifically. A slight disturbance in the background field.
I did not move. I held the layer absolutely steady and breathed in the slow, low-energy way the body had when it was minimizing output.
The entity completed its circuit and returned to the convergence point, forty meters from me.
It sat.
It spoke.
Not to me. Not to anything I could see. It spoke in a continuous low monologue that had the cadence of counting or cataloguing — repeated structures, periodic variations, the rhythm of something systematic rather than conversational.
I listened.
I could not understand a word.
But I could hear the structure. And structure was information even without content.
The language had three distinct registers — three sound frequency ranges that seemed to mark different categories of statement. The lowest register was used for the longest sequences, the most complex statements. The mid register was shorter, sharper. The high register was rare, used only twice in the twenty minutes I listened, each time followed by a pause.
Not a grammar I could decode in twenty minutes. But a shape.
I listened for four hours that day, memorizing sound patterns, storing them the way I stored everything — filed, organized, cross-referenced. When I left, I carried enough raw phonological data to begin the language design in the system with something above zero comprehension.
Not much above zero.
But above it.
[ DESIGN ASSESSMENT ]
Modification: Linguistic Comprehension — Infernal Tongue (Primary)Comprehension: Minimal (Phonological structure only; no vocabulary; no grammar)Quality: Very LowCost: 5 SoulsNote: This installs sound recognition and pattern storage, not comprehension. Host will be able to hear structural patterns more clearly but will not understand meaning. Comprehension grows through ongoing exposure. Reinstall not required — base linguistic modification will update as comprehension increases.
Five souls for a platform, not a result. The modification would install the hardware; understanding the language would come from time and exposure.
That was acceptable.
I approved it.
The change was subtle: the next time I heard any sound, I became aware that I was parsing it differently. Not understanding it — the advanced entity's language was still content-free to me. But I could hear the structure with more precision. The three registers were clearer, the boundary between sounds more distinct, the repetition patterns more legible.
A very long project. Started.
218 souls remaining.
I sat at the edge of the primary canyon in the late dark, running absorption, thinking about the anomaly in the southern wall. The portal. The energy of another world bleeding through.
I had been in the Infernal Realm for eighteen days. I had built, from a body that had arrived with nothing but baseline capability: four physical modifications, three energy modifications, one linguistic foundation. I had 218 souls and an expanded reservoir and a clarity about what the next thirty days needed to accomplish.
More souls. More comprehension. Closer study of the advanced entity. Language development. And eventually, when the mathematics were right, something I had been deliberately not looking at too directly, the way you don't look directly at a very bright light.
The thing below.
The source.
Whatever was generating the energy that the surface of this world fed from — whatever had built the convergence lines, whatever the advanced entity was ultimately drawing from — that was the question that mattered more than any other single question. Not for today. Not for a hundred days from now.
But the direction was down, and I had every intention of going there.
I watched the anomaly pulse in the southern wall. A slow, irregular breathing. A world on the other side, close enough to feel but not yet close enough to reach.
Eventually.
First: more.
Always more.
I stood up and went to find something to kill.