Anil-matcha/vox-ai-motion-graphics-generator
muapi-director
VerifiedTurn ONE topic, talking-head video, or photo into a finished Vox-style paper-collage explainer / ad video on the MuAPI platform (api.muapi.ai) + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Three input modalities: a topic (B-roll), a talking-head video (A-roll mode), or a single photo of a person/product anchored into the collage (C-roll mode).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.7 KB |
| assets/demo.mp4 | binary | 6.3 MB |
| assets/youtube-thumbnail.jpg | binary | 100.1 KB |
| examples/sample-beats.json | config | 2.4 KB |
| references/beat-layer.md | markdown | 3.3 KB |
| references/prompt-guide.md | markdown | 3.0 KB |
| references/voices.md | markdown | 1.4 KB |
| scripts/aroll_assemble.py | script | 2.6 KB |
| scripts/aroll_clips.py | script | 4.5 KB |
| scripts/asr_beats.py | script | 4.9 KB |
| scripts/assemble.py | script | 6.4 KB |
| scripts/audio.py | script | 3.7 KB |
| scripts/clips.py | script | 5.8 KB |
| scripts/croll_keyframes.py | script | 5.1 KB |
| scripts/keyframes.py | script | 3.1 KB |
| scripts/muapi_client.py | script | 9.0 KB |
| scripts/provider.py | script | 5.5 KB |
| scripts/style_bakeoff.py | script | 2.4 KB |
| scripts/styles.py | script | 11.5 KB |
| scripts/text_overlay.py | script | 4.6 KB |
scripts/clips.py
View on GitHub#!/usr/bin/env python3
"""
Clip stage: animate each SHOT's keyframe into a short video clip via MuAPI.
"""
import json
import os
import sys
from provider import get_provider, run_jobs
from styles import resolve_theme, resolve_video_aspect
VIDEO_MODEL = "veo3.1-image-to-video"
def shots_of(beat):
if beat.get("shots"):
for s in beat["shots"]:
yield s, f"{beat['id']}{s.get('id','')}"
else:
yield beat, f"{beat['id']}"
CAMERA_VOCAB = {
"static": "a locked-off static camera (no camera move)",
"push_in": "one very slow smooth push-in (uniform scale-up, Ken-Burns)",
"pull_out": "one slow smooth pull-out (uniform scale-down) revealing the full scene",
"pan": "one slow horizontal pan across the frame (flat translate, no perspective shift)",
"tilt": "one slow vertical tilt (flat translate, no perspective shift)",
"parallax": "a gentle multi-layer parallax drift (paper layers moving at slightly different speeds)",
"orbit": "a slow 3D orbit / camera arc around the scene",
"dolly_zoom": "a dolly-zoom — push in while the background pulls back",
"roll": "a slow camera roll",
"whip": "a fast whip-pan sweep that settles",
}
AMPLITUDE = {
"calm": "subtle, restrained amplitude",
"punchy": "lively, energetic amplitude with clear, bold movement",
"max": "high-energy amplitude — elements burst, scatter and fly boldly",
}
DEFAULT_ELEMENT = ("several cut-out elements move naturally with the scene — they bob, tilt, "
"slide, drift and scatter; halftone dots pulse")
def collage_prompt(camera, element_motion=None, feel=None, palette=None, amplitude="punchy",
has_title=True, constraints="strict", anchor_freeze=None):
cam = CAMERA_VOCAB.get(camera, camera)
element = element_motion or DEFAULT_ELEMENT
feel = feel or "tactile, editorial, a page from a scrapbook"
palette = palette or "the still's existing palette, high contrast"
amp = AMPLITUDE.get(amplitude, AMPLITUDE["punchy"])
text_lock = ("Keep the HEADLINE TEXT sharp, legible and stable — do not warp or wobble the lettering. " if has_title else "")
if anchor_freeze:
text_lock = f"{anchor_freeze} " + text_lock
if constraints == "loose":
guard = (text_lock + "Keep the paper-collage look (printed cut-outs, not photoreal); "
"avoid a subject melting into shapeless goo. Otherwise explore freely.")
else:
guard = (text_lock + "Keep the layout stable. Stay flat 2D — no 3D rotation, no "
"perspective change, camera parallel to the poster. ONE continuous move that "
"does not loop, retract or reset. Rigid paper — no morph/melt. Animate the "
"motion only; don't re-render the picture.")
return (
"Animate this still into a mixed-media paper-collage MOTION GRAPHIC, printed cut-outs, not photoreal.\n"
f"CAMERA (one move only): {cam}.\n"
f"ELEMENT MOTION (rich — this is the energy, {amp}): {element}. Elements move as paper cut-outs (slide, flap, hinge, pop, scatter, fly).\n"
"AESTHETIC: keep the torn-paper, tape, halftone, newsprint and paper-stencil textures and the bold flat background.\n"
f"FEEL: {feel}.\nCOLOR: {palette}.\n"
f"CONSTRAINTS: {guard}"
)
def run(project_dir):
bpath = os.path.join(project_dir, "beats.json")
with open(bpath) as f:
doc = json.load(f)
model = doc.get("video_model", VIDEO_MODEL)
target_aspect = doc.get("aspect", "16:9")
resolved_aspect, exact = resolve_video_aspect(target_aspect, model)
theme = resolve_theme(doc.get("theme")) or {}
t_amp = theme.get("motion_style", "punchy")
clip_dir = os.path.join(project_dir, "clips")
os.makedirs(clip_dir, exist_ok=True)
prov = get_provider(doc.get("provider"))
specs, by_key = {}, {}
for beat in doc["beats"]:
b_feel = beat.get("feel")
for shot, key in shots_of(beat):
if shot.get("clip_url") and os.path.exists(shot.get("clip_path", "")):
continue
kf = shot.get("keyframe_url")
if not kf and shot.get("keyframe_path"):
kf = prov.upload(shot["keyframe_path"])
shot["keyframe_url"] = kf
if not kf:
print(f"[{key}] no keyframe_url, skipping")
continue
cam = shot.get("camera_move", "push_in")
elem = shot.get("element_motion")
amp = shot.get("motion_style") or t_amp
constr = shot.get("constraints", "strict")
prompt = collage_prompt(cam, elem, feel=b_feel, amplitude=amp,
has_title=shot.get("title", True), constraints=constr,
anchor_freeze=beat.get("anchor_freeze"))
shot["clip_prompt"] = prompt
dur = 8 if "veo3" in model else int(shot.get("duration", 5))
params = {"image_url": kf, "duration": dur}
if resolved_aspect:
params["aspect_ratio"] = resolved_aspect
specs[key] = (lambda p=prompt, par=params: prov.submit_video(model, p, **par))
by_key[key] = shot
done = run_jobs(prov, specs, poll_s=4, stall_s=120, max_retries=2, deadline_s=900)
for key, url in done.items():
if not url:
continue
dest = os.path.join(clip_dir, f"clip_{key}.mp4")
prov.download(url, dest)
shot = by_key[key]
shot["clip_url"] = url
shot["clip_path"] = dest
print(f"[{key}] saved {dest}")
with open(bpath, "w") as f:
json.dump(doc, f, ensure_ascii=False, indent=2)
print("Updated", bpath)
if __name__ == "__main__":
proj = sys.argv[1] if len(sys.argv) > 1 else os.path.join(
os.path.dirname(__file__), "..", "out", "demo")
run(os.path.abspath(proj))
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |