The AI-native design pipeline — concept to shipped MP4 in one workflow — step 7 of 8

Write score_pipeline(steps) that takes a list of step dicts and returns an integer 0-100.

• Canonical step names: ["concept", "reference-gen", "composition", "audio", "captions", "render", "review"]
• For each entry in steps, count it as "covered" if BOTH:
  • its step field is in the canonical list, AND
  • its tool field is a non-empty string.
• Each unique canonical step covered = 15 points.
• Cap the score at 100.
• Return an int.

Three teams run. Expected output:

team alpha: 100
team beta:  90
team gamma: 45