This is the third time I've seen everyone get carried away by "testnet incentives/points"… I won't oppose the hype, after all, hype is just traffic, but don't mistake traffic for assets. To be honest, first ask yourself two questions: Is there a verifiable return path for the actions you take (not guesses about whether the mainnet will issue tokens), and is there a cap on the time/fees you invest? Many projects just lock people into tasks with points, and when the mainnet comes and rules change, the story shifts, but the data remains—the active addresses, retention, real transaction volume—spread out, and the hype instantly cools down. My approach is pretty simple: only do reusable actions (like automating processes, recording costs), stop once the budget is exceeded. Better to miss out than to be repeatedly cut by the same script.