Pose Test — flux2-klein image-edit

Identity from the reference image · pose driven by TEXT instruction (no target-pose image yet)

IDENTITY REF (input) cassidyfresh_00

The source image. Identity (face + body) is supposed to carry from here.

RESULT (output) pose_f2edit_A

flux2-klein edit. Pose instruction: "standing, one hand on hip, weight on one leg, three-quarter turn." Vision check: clean body, correct anatomy. YOU judge the face/identity hold.

Note: the "pose" here was a TEXT instruction, not a target-pose image. For a true pose-to-pose compare, next run uses an actual target-pose reference so you can see [target pose] vs [result] side by side.