library/moments.jsonl (precedent rows)

Agent skills · A plan or a script · Apache-2.0

What it does

716 rows of rule-only precedent (one more line is a header): 507 rules decoded from other human-made films under opaque ids, 156 designer rules written as when/do/because with numbers, and 53 rows for the measured study moments. No copy, frames or brand names. The file's own header says the rows are model-decoded and unverified.

Get it

python3 scripts/precedent.py query --beat proof --content ui --text "selection becomes result" --k 6 (also: show ID, rules --beat cta, stats)

Source on GitHub
Needs (the whole pack)

Node 22, Python 3.9, FFmpeg and Chrome. Optional and paid: an ElevenLabs key for voice and music. No Remotion: a film is an HTML page rendered through Chrome.

License

Apache-2.0

License notes are on the pack's page

From

motionmaxxing

Everything from motionmaxxing in the gallery

Back to the Asset Library