TimoteoAltobrando/ai-video-skill
ai-video
UnverifiedMake and edit videos with Claude Opus 5.5, end to end, from a blank page or from the user's own footage — product and launch films, ads, talking-head and founder videos, interviews, explainers, tutorials and screen demos, event and trip montages, highlight reels, podcast clips, Reels/TikTok/Shorts, animated logos and intros, music and lyric videos, and changes to a finished video (shorter cut, vertical version, subtitles, translation, new music). Remotion is the backbone; Codex (image_gen) makes images, Seedance via OpenRouter animates them, ElevenLabs makes music and sound effects, ffmpeg and whisper handle footage and speech. It starts by asking what kind of video it is and everything that kind needs (platform, format, voice, music, rights, who must not appear, budget), then goes through direction, a script or paper edit approved piece by piece, production, revisions and delivery for each platform. Use it whenever someone wants to make, cut, edit, shorten, reformat, subtitle or improve a video, even if they do not mention any of these tools.
scripts/inventory.py
View on GitHub#!/usr/bin/env python3
"""Inventory of a folder of footage: one row per clip, and optionally one contact sheet per clip.
Usage:
python3 inventory.py footage/original --out docs/footage.md [--sheets docs/sheets] [--every 5] [--scenes 0.35]
It reads every video file (mov, mp4, m4v, mkv, webm, avi, mts) with ffprobe and writes a markdown table: file, duration,
size, frame rate (and VFR when the rate is variable, common on phones), codec, HDR (HLG / PQ), rotation, audio, date.
--sheets writes <name>.jpg for each clip: one frame every --every seconds, in a grid with the time in the corner.
Look at them before choosing anything: they are how you "watch" hours of footage.
--scenes also counts the cuts or big changes inside each clip (ffmpeg scene score above the threshold); useful to
find the shots inside a long recording that was already edited.
The originals are only read, never changed.
Needs ffmpeg/ffprobe, and Pillow for --sheets.
"""
import argparse
import json
import pathlib
import subprocess
import tempfile
EXT = {".mov", ".mp4", ".m4v", ".mkv", ".webm", ".avi", ".mts"}
def probe(p: pathlib.Path) -> dict:
r = subprocess.run(["ffprobe", "-v", "error", "-show_format", "-show_streams", "-of", "json", str(p)],
capture_output=True, text=True)
return json.loads(r.stdout or "{}")
def rate(s: str) -> float:
n, _, d = (s or "0/1").partition("/")
return float(n) / float(d or 1) if float(d or 1) else 0.0
def rotation(v: dict) -> int:
for sd in v.get("side_data_list") or []:
if "rotation" in sd:
return int(sd["rotation"])
return int((v.get("tags") or {}).get("rotate", 0))
def scenes(p: pathlib.Path, threshold: float) -> int:
r = subprocess.run(["ffmpeg", "-hide_banner", "-nostats", "-i", str(p), "-vf",
f"scale=320:-2,select='gt(scene,{threshold})',showinfo", "-an", "-f", "null", "-"],
capture_output=True, text=True)
return r.stderr.count("pts_time:")
def sheet(p: pathlib.Path, out: pathlib.Path, every: float) -> None:
from PIL import Image, ImageDraw
with tempfile.TemporaryDirectory() as tmp:
subprocess.run(["ffmpeg", "-v", "error", "-i", str(p), "-vf", f"fps=1/{every},scale=320:-2",
f"{tmp}/f%05d.png"], check=True)
frames = sorted(pathlib.Path(tmp).glob("f*.png"))
if not frames:
return
w, h = Image.open(frames[0]).size
cols = min(6, len(frames))
rows = (len(frames) + cols - 1) // cols
grid = Image.new("RGB", (cols * (w + 4) + 4, rows * (h + 4) + 4), "black")
for i, f in enumerate(frames):
im = Image.open(f).convert("RGB")
t = int(i * every)
ImageDraw.Draw(im).text((6, 4), f"{t // 3600}:{t // 60 % 60:02d}:{t % 60:02d}", fill="white",
stroke_width=2, stroke_fill="black")
grid.paste(im, (4 + (i % cols) * (w + 4), 4 + (i // cols) * (h + 4)))
grid.save(out, quality=85)
def main() -> None:
a = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
a.add_argument("folder")
a.add_argument("--out", required=True)
a.add_argument("--sheets")
a.add_argument("--every", type=float, default=5.0)
a.add_argument("--scenes", type=float)
x = a.parse_args()
files = sorted(f for f in pathlib.Path(x.folder).rglob("*") if f.suffix.lower() in EXT and f.is_file())
if not files:
raise SystemExit(f"no video files in {x.folder}")
if x.sheets:
pathlib.Path(x.sheets).mkdir(parents=True, exist_ok=True)
head = "| File | Duration | Size | fps | Codec | HDR | Rotation | Audio | Date |" + (" Cuts |" if x.scenes else "")
rows = [head, "|" + "---|" * (head.count("|") - 1)]
total = 0.0
for f in files:
info = probe(f)
v = next((s for s in info.get("streams", []) if s.get("codec_type") == "video"), None)
if not v:
continue
au = next((s for s in info.get("streams", []) if s.get("codec_type") == "audio"), None)
dur = float(info.get("format", {}).get("duration", 0))
total += dur
fr, avg = rate(v.get("r_frame_rate")), rate(v.get("avg_frame_rate"))
fps = f"{avg:.2f}" + (" VFR" if fr and avg and abs(fr - avg) > 0.05 else "")
trc = v.get("color_transfer", "")
hdr = {"arib-std-b67": "HLG", "smpte2084": "PQ"}.get(trc, "no")
date = ((info.get("format", {}).get("tags") or {}).get("creation_time", "") or "")[:16].replace("T", " ")
audio = f"{au.get('codec_name')} {au.get('channels')}ch" if au else "none"
m, s = divmod(dur, 60)
row = (f"| {f.relative_to(x.folder)} | {int(m)}:{s:04.1f} | {v.get('width')}×{v.get('height')} | {fps} | "
f"{v.get('codec_name')} {v.get('pix_fmt', '')} | {hdr} | {rotation(v)} | {audio} | {date} |")
if x.scenes:
row += f" {scenes(f, x.scenes)} |"
rows.append(row)
if x.sheets:
sheet(f, pathlib.Path(x.sheets) / f"{f.stem}.jpg", x.every)
print(f"ok {f.name}")
m, s = divmod(total, 60)
rows += ["", f"{len(rows) - 2} clips, {int(m // 60)} h {int(m % 60)} min {int(s)} s in total."]
pathlib.Path(x.out).parent.mkdir(parents=True, exist_ok=True)
pathlib.Path(x.out).write_text("\n".join(rows) + "\n")
print(f"-> {x.out}")
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |