{"id":"vector-normalize","version":"1.0.0","description":"Scale a finite-number vector to unit Euclidean length.","supported_operations":["vector normalization","unit vector","make vector unit length"],"unsupported_operations":["normalizing the zero vector (undefined direction)","normalization by a norm other than L2"],"semantics":["Input vector is an array of 1 to 64 finite JSON numbers, each bounded to at most 1000000 in magnitude.","The Euclidean (L2) norm is computed with IEEE-754 double-precision Math.sqrt over the sum of squares.","Optional input tolerance (default 1e-9, range 0 to 1) is the minimum norm accepted; vectors whose norm is at or below tolerance are rejected as having no well-defined direction.","The result is result[i] = vector[i] / norm, computed in IEEE-754 double precision; -0 is normalized to 0.","Every result entry is checked to be a finite number before it is returned; a non-finite intermediate is rejected as invalid input rather than silently serialized."],"limits":{"max_dim":64,"max_abs_value":1000000},"pricing":{"status":"unpriced","charge_usd":null},"input_schema":{"type":"object","additionalProperties":false,"required":["vector"],"properties":{"vector":{"type":"array","minItems":1,"maxItems":64,"items":{"type":"number"}},"tolerance":{"type":"number","minimum":0,"maximum":1}}},"output_schema":{"type":"object","additionalProperties":false,"required":["result"],"properties":{"result":{"type":"array","items":{"type":"number"}}}},"examples":[{"input":{"vector":[3,4]},"output":{"result":[0.6,0.8]}}],"execute_url":"/v1/tools/vector-normalize/versions/1.0.0/execute"}