Start with a role, not a tier list
The official store describes over 200 skills and gives examples of powerful attacks, defense and healing, and gradual damage through poison. These examples confirm different strategic directions; they do not establish a ranked list of best skills.
For each skill available in your run, read the current description and identify its main role. Decide whether it produces damage, helps the dragon survive, or improves another part of the plan. Do not rank a skill solely by a large-looking number.
Look for connected effects
A useful combination has a clear reason for each choice. If one skill supports another effect, check the wording and conditions in the game rather than assuming they interact. Distinguish a combination you have tested from one that only appears promising.
Try to keep your comparison simple. Change a small number of related choices, then compare a similar encounter. Several simultaneous changes make it difficult to tell what helped.
Diagnose why a battle failed
If the dragon is defeated before the plan develops, investigate survival. If the battle lasts but the enemy is not overcome, review how the damage plan functions. If a desired effect never appears, reread its requirements and observe when it activates.
These are troubleshooting questions, not claims about hidden enemy defenses, poison formulas or mandatory skill combinations. We have not independently verified those numerical mechanics.
Keep version and evidence together
Record the game version, chosen skills and circumstances when you find a useful combination. Balance updates can change how advice performs. A guide should explain its assumptions and evidence before calling a build optimal.
This page deliberately omits unsupported skill costs, tier rankings and boss-specific guarantees. The training guide explains how to connect your growth decisions to a combat direction.