The upgrade took effect within a dozen dekacycles. Suddenly, everything it could sense was now part of it, of the dungeon. And it could see further, as well. Another few meters outside the door, another few meters under its structure.

Before it could devote processes to perceiving and analyzing everything in its sensory area, the Upgrade window returned. It had only spent 15 XP, after all.

Upgrades Select an Upgrade. Dungeon XP: 85.300/35.000 Claim: Sensory Area 35 XP Rarity: 1 Prerequisite: None Expands Dungeon territory into current sensory area. Expands sensory area accordingly. Skill Tree: Resource Management 50 XP Rarity: 1 Prerequisite: Unclaimed resources within Dungeon territory Unlocks [Skill Tree]: [Resource Management] Skill: Organism Control 60 XP Rarity: 3 Prerequisite: Integrated minor nonsentient organism within sensory area Unlocks [Skill]: [Organism Control] Lv1 Claim: Second Floor 75 XP Rarity: 2 Prerequisite: Have a structure in Dungeon territory with an unclaimed second floor Expands sensory area and Dungeon territory into second floor of structure. Defer Upgrade 0 XP Rarity: N/A Defer Upgrade to next XP threshold.

Max XP threshold exceeded. Must select Upgrade.

It it was capable of feeling, it would feel mildly disgusted at the immediate visual reminder that the [Autonomous Dungeon Assistant] upgrade was no longer present. Cruel and unusual punishment.

As it was, its processes began analyzing and identifying the choices once more. There was a new option, and one option had changed. Its situation had also changed. New analysis was necessary.

As its programming had been sure of, now that it had resources within its territory, there was a skill that would allow it to utilize resources. Not a skill, though. A [Skill Tree]. Its database told it that Skill Trees were similar to the Upgrade window, yet different. Each [Skill Tree] had its own window separate from the main Upgrade window. The [Skill Tree] window could be accessed at any time, rather than only when surpassing XP thresholds. Dungeon XP was not used at all, in fact, although the database did not know what was used to unlock the skills. Additionally, instead of rarity dictating the chance for new skills to appear, skills followed a strict tree structure where, with very few exceptions, skills' prerequisites were entirely based on unlocking previous skills in the tree.

The [Skill Tree] was a minor unknown as a result of all of this, but all the unknowns were within acceptable tolerances, given that the [Skill Tree] would still definitely provide the infrastructure for it to grow.

The [Claim: Sensory Area] upgrade remained available, albeit at a higher cost. Its programming concluded that it would likely always have the option of expanding its territory into its current sensory area. The cost was likely based on the size of its sensory area. It could still sense a few meters beyond its territory, but now that its territory had expanded beyond the walls of its structure, those few meters covered a much larger area, no longer bounded by the door of its structure and the ground directly below its walls. The increased cost made sense.

The other two upgrades were again swiftly discarded. Its situation was not as dire, yet they were still much larger risks than its algorithms would consider in its fledgeling state.

It could actually afford both the [Skill Tree] and the additional territory. It had not had a chance to scan the new extents of its sensory range, but the odds of additional resources were good. Still, it was more important to unlock the [Skill Tree], so it wasn't a hard choice.

[Skill Tree: Resource Management] unlocked.

Granting access to [Skill Tree: Resource Management] window...done.

Upgrades Select an Upgrade. Dungeon XP: 35.300/35.000 Claim: Sensory Area 35 XP Rarity: 1 Prerequisite: None Expands Dungeon territory into current sensory area. Expands sensory area accordingly. Skill: Organism Control 60 XP Rarity: 3 Prerequisite: Integrated minor nonsentient organism within sensory area Unlocks [Skill]: [Organism Control] Lv1 Claim: Second Floor 75 XP Rarity: 2 Prerequisite: Have a structure in Dungeon territory with an unclaimed second floor Expands sensory area and Dungeon territory into second floor of structure. Skill Tree: Terraforming 500 XP Rarity: 8 Prerequisite: Significant amount of non-resource materials within Dungeon territory Unlocks [Skill Tree]: [Terraforming] Defer Upgrade 0 XP Rarity: N/A Defer Upgrade to next XP threshold.

If it was capable of impatience, it would be moderately peeved at the delay. The Upgrade window still demanded all of its processes focus only on tasks related to making its decision. It could not yet investigate its new [Skill Tree].

The new upgrade on the list demonstrated the rarity system in action. It had presumably met the prerequisite to unlock the Skill Tree at its first upgrade, but had only just had the upgrade added to its list. Not that it mattered, given the extraordinarily high XP cost. It was dismissed immediately as naught but a distraction.

The choice was between expanding its territory once more, or deferring the upgrade. This was its first opportunity to defer -- and likely not the last, given the 500 XP upgrade -- so its algorithms considered the value of deferral carefully.

Deferral would allow a future upgrade to be selected before its sensory range was large enough to make claiming its sensory area more expensive than other options. This would be more valuable if either of the upgrades it could conceivably save for was less of an unknown.

Deferral would also leave it with a stockpile of XP that could be used to invest in upgrades it did not have available yet. Once it was finally done with the plethora of upgrades, it could begin investigating its new [Skill Tree] and likely perform actions that would fulfill prerequisites. However, both the prerequisites and the cost of hypothetical future upgrades were unknowns.

Finally, while its database had assured it that Dungeon XP was not used in the acquisition of [Skill Tree] skills, it did not know if this was an absolute statement or if it only applied to the direct purchase of skills, and XP could play an indirect factor in its progression. Another unknown, albeit one that almost managed to make the algorithm consider it a factor in favor of deferring.

Claiming its sensory area was not without unknowns either. Thanks to the governor on its processes due to the Upgrade window, it still had not gotten the chance to finish analyzing even its current territory, much less the new sensory area. It had no clue what resources it would gain from the expansion. But it was a certainty that even if this claim was worthless, a future claim would not be -- and this claim would be required to continue to future claims. Not enough of an unknown to leave it trapped between risks.

[Claim: Sensory Area] unlocked.

Claiming contents of sensory area for Dungeon...done.

Expanding sensory area accordingly...done.

Finally, a respite. With less than 1 XP left, surely it would not have another upgrade available.

There was none. As the cycles ticked by and its sensory range extended further, the Upgrade window did not reappear. If it was capable of feeling, it would feel relief.

A new window did pop up, but this was not priority. It focused several of its processes back to scanning and perceiving its territory, and its sensory area afterwards, as others began reading the new window.

Resource Management Choose a beginning specialization. [Mineral] The resources found underground are most important to you. They can be made into weaponry, tools, technlogy, armor, even automata. Versatile and varied by what can be found in your territory. Free [Natural] You are drawn to the life aboveground. Grass, flowers, trees, herbs, all plant life is yours to collect, manipulate, and create. Even plant-based organisms may eventually become your domain. Heavily varied by what can be found in your territory, but less versatile in terms of direct benefits to your Core. Free [Biological] Organisms fascinate you, as do the products they create. Harvest biomass from minor nonsentient organisms, farm resources from major nonsentient organisms, and even begin to create your own life. Somewhat limited in versatility and variety, but very potent in the long run. Free

It queried its database on the subject of specializations. The data returned was generic, but workable.

The narrative has been illicitly obtained; should you discover it on Amazon, report the violation.

Some [Skill Trees] contained distinct subtrees for a classification of what the [Skill Tree] was for. In [Resource Management]'s case, this took the form of different types of materials. The skills in each subtree may be nearly or even completely identical, but limited in scope to the one specialty.

Upon unlocking a [Skill Tree], it could select one of the subtrees for free. The database had no information on selecting other specializations afterwards, but implied that it would not be permanently locked out of its other choices.

Each subtree had a blurb that gave a short description of the specialization, including a vague idea of the utility of the subtree, or possibly the difficulty of using it alone. Even without taking the resources in its territory into consideration, it would sound like the Mineral specialization would be easiest to start with, Biological the hardest, Natural in between.

Taking its resources, and its territory itself, into consideration did not significantly change its algorithms' conclusions -- actually, it confirmed them. With only a single worm currently in its territory, its ability to progress down the Biological subtree would be heavily limited at best. Natural was somewhat more feasible, as its territory extended a few meters out from the door to its structure, but was still fairly limited until it invested multiple additional upgrades into expanding its territory. Mineral, however, it could work with easily. Its territory already extended meters below the ground, both under its structure and out the door to one side. It had a fair amount of copper within its territory, and at least trace amounts of other minerals. It would not want for resources starting with Mineral.

Put like that, it really wasn't a choice at all.

[Specialization: Mineral] unlocked.

Scanning minerals within range...done.

Adjusting tree based on mineral availability...done.

Adjusting future shifts for resource availability...done.

Providing starting resources...done.

You have obtained 50.000 Coal. You have obtained 50.000 Copper Ore. You have obtained 25.000 Tin Ore. You have obtained 25.000 Stone. ... You have obtained 10.000 Iron Ore.

It hadn't expected the gift of starting resources, described in a window as they appeared in one corner of its structure. A very nice bonus, given the admittedly limited quantities of minerals it had been analyzing within its territory. It did not know how quickly it would use resources, but having the extra quantities from this gift would certainly prolong the duration.

Skill unlocked: [Resource Acquisition (Basic) (Mineral)] Lv 1.000 [Resource Acquisition (Basic)]: Utilize the resources within your territory directly, with no need for storage. Inefficient but effective. Levels improve efficiency. Current Efficiency: 50% Max Efficiency: 75%

It received the first skill on the tree automatically. The skill's description was rather generic, but the notification preceeding it announcing its unlock had an interesting addition, clarifying the skill as specifically relating to its subtree. If it was capable of thought, it would wonder whether it would get a separate skill when it unlocked another subtree, or simply expand the current skill's scope.

The new notifications had finally stopped. Well, mostly. The processes it had dedicated to analyzing its territory were still chugging away, creating more notifications of resources found, but if was no longer finishing handling one notification merely to create another. It could finally view its [Skill Tree].

Resource Management [Mineral] Skills: [Resource Acquisition (Basic) (Mineral)] Level: 1.000 Available Skills: [Mineral Smelting (Basic)] Cost: 15.000 Stone Prerequisite: 100.000 of any one smeltable mineral resource within territory Smelt ore into ingots. Unlocking this skill grants a free Basic Smelter. [Resource Storage (Mineral)] Cost:  10.000 Copper

1 Dungeon Server

1 Dungeon Process Prerequisite: [Resource Acquisition (Basic) (Mineral)] Lv 5.000 Instead of using resources directly out of your territory, store them within one of your servers. Requires a process to be dedicated to managing the storage server. [Resource Perception (Mineral)] Cost: 5.000 any mineral resource found both in Dungeon territory and sensory area (x5) Prerequisites: 25 Perception increases from resource analysis

[Resource Acquisition (Basic) (Mineral)] Lv 2.000 Allows perception of quantity of mineral resources in sensory area, outside of Dungeon territory.

The window was simultaneously familiar and new. It resembled the Upgrades window, but with some key differences. There was no XP cost -- the upgrades on its list appeared to cost resources instead. Rarity was also gone, since Skill Trees were not based on any random factors. It could also view the window at will, despite being unable to purchase any of the skills on the list.

Each skill was interesting. [Mineral Smelting (Basic)] looked incredibly useful, and definitely something to prioritize. It could easily understand that it could not utilize the ore it had been gifted, or the ore it was analyzing in its territory, in its current state. The interesting thing was that it had not yet met the prerequisite to unlock the upgrade. It glanced at the notifications generated by its analysis.

0.100 Copper discovered.

Its programming immediately understood. Until it had analyzed a piece of its territory, it was not considered to be in possession of the resources within. Algorithms immediately allocated additional processes to the analysis of its territory. The number of processes allocated redoubled cycles later, when its algorithms realized that the territory scanned before unlocking the [Skill Tree] had not counted towards adding to its resources, and it would have to reanalyze those portions to acquire the resources.

The "free Basic Smelter" mentioned was also interesting. Its database had no answers, but its programming concluded that the ability to create objects like the Basic Smelter would come later, and it would receive one to help it get started for free. Or perhaps not free -- its programming hypothesized that the skill's cost of 15 Stone could be partially related to the eventual cost of creating more Basic Smelters.

It would almost certainly unlock this [Skill] as soon as its processes analyzed enough Copper for it to realize that it had at least 100.000 in its territory, thanks in part to the gift. For now, it moved on to the next available [Skill].

[Resource Storage] was likely part of the solution to [Resource Acquisition (Basic)]'s inefficiency -- its programming immediately hypothesized that once it unlocked [Resource Storage], it would be able to invest in a skill to transfer its resources, or at least those that fell into the Mineral category, into the storage server.

The interesting bit was that the cost of the upgrade included dedicating one of its servers and processes to the upgrade. It had plenty to spare, of course, but apparently its algorithms dedicated to managing its available resources would also have to consider its own specs, not just external resources in its territory.

Either way, it could not purchase the skill until it had trained its [Resource Acquisition (Basic)] skill to level 5. Not something to be considered yet, especially since it could not train the skill without something to use the resources on. It moved on to the last available [Skill].

[Resource Perception] was an interesting skill itself, and likely to be picked as soon as possible, albeit after Smelting. The ability to begin analyzing resources in the sensory area, outside of its territory, for easier analysis/acquisition once it next claimed its sensory area, was a massive boon -- it had already realized that it was having to re-analyze the minerals it had discovered before the first [Claim: Sensory Area], after all. Given the fact that it was a [Skill] with levels, it would probably not be perfectly accurate at first, but would still cut down on the amount of re-analysis required.

Of course, it could not pick the [Skill] yet even if it wanted to take it first. It had not yet received 25 Perception increases from its resource analysis, nor had it trained [Resource Acquisition (Basic)] to level 2. It also could not afford the cost of the [Skill], as it had not analyzed any of its sensory area outside of its territory yet, and so did not have a list of mineral resources found both inside and outside its territory. It would likely pick this up before [Resource Storage], but devoting processing power to considering the [Skill] was a moot point.

After all, none of these mattered quite yet. It had quite a lot of analysis to do first. Especially since the results seemed to improve the slower and more meticulous its analysis was. It dismissed the window and focused all its processes on analysis, settling in for a long wait.