I'll make a list. Note that not all of them apply at the same time, it often depends on the idea:

• if it's concrete and simple enough, I may just test it out, experiment with it
• meditate on it, especially if it's not very concrete, but rather blurry
• research it (nowadays I may use AI too for that)
• make some calculations to see if it's feasible or if there are multiple options what's the best choice (again, I may use AI)
• if I need advice or a second opinion, talk to someone, either in real life or online; it may be here on Hive if it applies; I may write a post or be engaged in a discussion about the idea