Learn
Not a product tour. Specific techniques, ready-made prompt fragments and mistakes we walked into ourselves. No account needed to read any of it.
The skills don't make videos, and neither does Claude on its own. The video is made by a video generator, such as Higgsfield or Kling, on your own subscription. Here is every step, from the idea to the finished film.
Read itGenerators default to flawless skin, perfect symmetry and model proportions. That is how something meant to be a drama comes out looking like a perfume ad. Four techniques for getting a human face instead.
Read itThe most common failure in AI video: by the second scene, your lead is somebody else. The fix isn't a longer prompt. It is not working from text at all.
Read it„Cinematic shot" is a mood, not an instruction. A filmic image is six independent decisions, and the generator expects each of them on its own.
Read itWhat you don't ask for matters as much as what you do. A well-built blocklist improves an image more than twenty further adjectives.
Read itThe image is handsome, the music is right, and still nobody watches to the end. Usually because everything is where it started. Five questions that catch it.
Read itTime of day, direction, quality, tone: four clicks build a lighting line the model cannot misread, and you can copy the finished prompt straight away.
Read itHands are the most common flaw in AI images, and they fail for a reason. Say what the hand is doing, what it holds and which fingers show, and the generator stops guessing.
Read itIf your film has a person, a product or shots that must match, start from an image. Here is why, and when plain text is still enough.
Read itAsk a video model for a slow push-in and you often get a frozen shot. Here is how to write a camera move as distance, time, and a first and last frame.
Read itEverything you read here exists inside the system as something you click, from the blocklists to the camera angles. You don't have to memorize it.