Day 114.
Assessment Day.
I had eaten breakfast alone, which was a deliberate choice. Maris had understood without asking. Some things you needed to walk into with your own thoughts and no one else's.
The written examination was in the morning. Three hours, covering mana theory, cultivation methodology, combat application principles, and — my personal favorite — applied dungeon scenario analysis. I had been dreading the last one for three months.
Not because I didn't know the answers.
Because I knew them too well.
Knowing too much was, I had discovered, its own category of problem.
------------------------------------
The examination hall seated all Class A and B first-years simultaneously. Three hundred students, twelve rows, the specific quality of enforced silence that descended over any room full of anxious people who had been told to be quiet and work.
I found my assigned seat. I had a system.
For standard theory questions — mana circulation mechanics, affinity classification frameworks, core formation typology — I answered at the level of a well-read first-year. Accurate. Organized. Exactly what the rubric rewarded. I had checked the standard curriculum for each topic.
For technique application questions — the ones asking about optimal mana usage in specific combat scenarios — I let myself go a little further. Not to the graduate-level depth I could have reached. To the level of someone who had been obsessively practicing those techniques in a basement for four months. Defensible.
The problem was question seventeen.
Question seventeen read: "An SS-ranked void entity has forced a partial mana-grid locus breach at a Foundation-era architectural site. Describe the characteristics of the resulting environmental anomaly and outline the principles a practitioner team should apply in a rapid-response engagement."
I stared at this question for a full sixty seconds.
This question.
The answer was in my notebook. Not because I had studied for it — because I had been living it for three months. I had documentation on void-mana grid infusion techniques, locus breach threshold mechanics, Foundation architecture interaction with rift-adjacent entities, and the precise limitations of rapid-response engagement against SS-ranked void presence.
I could write twelve pages on this question from primary observation.
I wrote one and a half. I focused on the theoretical principles, kept my response at the level of an advanced student who had read the right textbooks. I did not, for example, mention that I personally knew the specific grid juncture in the academy's basement where a cult operative was currently performing exactly this operation.
The invigilator announced thirty minutes remaining.
I finished question twenty-three, closed my examination paper, and spent the remaining time reviewing my answers for any place where I had accidentally been too specific.
There were two. I edited them down.
------------------------------------
The combat chamber assessment was the afternoon session.
Each student had twenty minutes in the chamber with a series of progressively more complex golem constructs. The chamber's evaluation system tracked technique output, mana efficiency, control precision, combat decision-making, and something the rubric called "situational reading" — the practitioner's ability to gather information from the fight environment and apply it in real time.
I stood outside the chamber door and ran a quick self-check.
Mana pool: 640. Full.
Ambient draw: active. The air in the assessment building's corridor was clean and circulating. I could feel it.
Draft Reading: passive, fifteen-meter range. Ready.
Narrative Sense: still dormant. Couldn't count on it.
Instructor Brennan was at the assessment console with a clipboard, watching the student currently inside finish their session. He was a compact man in his late forties who taught Advanced Combat Theory with the specific precision of someone who had spent twenty years identifying exactly what students were and weren't doing in fights.
He was also, I had gathered over the semester, the kind of instructor who updated his rubric mid-session when a student did something he hadn't seen before.
The door opened. The previous student walked out, sweat-damp and grimacing at whatever their score had been. Brennan made a note. His eyes moved to the next name on his list.
He looked up at me.
"Martin. In."
I went in.
------
The chamber's first golem was a standard Class B construct — squat, single-limbed strike pattern, predictable attack sequences designed to test basic technique application. Most students used this one to warm up their mana output.
I used it to calibrate the room.
Draft Reading pulled the air currents from wall to wall in 0.3 seconds. The chamber was twelve meters by twelve meters, sealed except for two ventilation grates. The golem's movement displaced air in a predictable pattern — the same way any solid moving object displaced air, which meant that Draft Reading, at my current level, effectively gave me a real-time map of the golem's exact position without needing to look at it.
I could fight it with my eyes closed.
I didn't, obviously. That would have been dramatic in the wrong direction.
The golem charged. I was already three steps to its left — Wind Step burst, quiet, efficient — and the Cutting Edge was along my forearm at half-density. I didn't need full density for the first construct. I used enough to register a clean hit on the scoring crystal.
Brennan made a note outside the glass.
The second golem came out. Faster. Dual-strike configuration. This one was calibrated for Class A students.
I activated Draft Reading at full output.
The ambient draw kicked in from the ventilation grates. The range extended to twenty-two meters — wall to wall, then through the ceiling grates into the support structure above. I felt the golem's movement before it committed to the strike pattern. Felt the air displacement from its shoulder joint as it cocked back.
I had moved before the strike began.
Not by much. Not dramatically. Just enough that the golem's first strike passed through air and its second strike found a Pressure Shell instead of an unprotected torso. The Shell absorbed sixty percent of the impact. I redirected the remaining momentum with a Wind Step angled to put me behind the golem's center line.
Cutting Edge.
The golem's scoring crystal lit amber.
Outside the glass, Instructor Brennan had stopped writing.
He was looking at me.
I kept going.
The third construct was a chimera configuration — multiple strike vectors, mana-discharge capability, the kind of enemy that punished practitioners who relied on a single strategy. I had been waiting for this one. It was the construct the assessment rubric used as the Class A differentiation filter: most first-years either retreated into pure defense or burned their mana pool on overwhelm attacks.
I did neither.
The ambient draw was running hot now — the circulating air from three active constructs' movement patterns feeding into my channels in a steady trickle. The mana pool read 590. I had spent fifty units on ten minutes of engagement. At standard output without ambient draw, I would have spent one-fifty.
I let the chimera come.
The first discharge — mana-energy bolt, center-mass targeting — I read two steps in advance through the air displacement shift and simply wasn't there. The second discharge I read one step in advance and deflected with Pressure Shell's edge rather than absorbing it full-on. The Shell cracked. I dismissed it immediately and rebuilt it at twenty percent density — enough to blunt a follow-up, not enough to waste mana on full protection.
Then I moved in.
Close range. Wind Step at full output from the left, cutting angle to the chimera's weak joint. Cutting Edge at full density. One clean strike to the scoring crystal.
The chimera froze.
Assessment complete.
I stood in the middle of the chamber, breathing at one-twenty beats per minute, and heard Instructor Brennan do something outside the glass.
He updated my score. Then he updated it again.
Then he called the next student.
_________