datit309/supergraph / plugins/supergraph/skills/video-shotcraft
video-shotcraft
UnverifiedCreate cinematic product videos from shot recipe cards, a validated template, and code/audio assets (Remotion + real page screenshots + 2.5D camera moves + beat-synced cuts + sound design). Use when the user asks to turn a frontend project or webpage into a product video, says "use video-shotcraft to make a video/promo", names the Ink Press template or asks to reproduce its effect, or wants a single shot card's motion. 用镜头配方卡 + 已验收模板 + 代码/音频资产制作电影感产品视频(Remotion + 真实页面截图 + 2.5D 运镜 + 节奏卡点 + 声音设计)。当用户要求"用 video-shotcraft 做视频/宣传片"、把前端项目/网页做成产品视频、点名 Ink Press 模板或要求复刻模板片效果,或要用镜头卡做单个动效镜头时使用。
.github/scripts/showcase-publish.py
View on GitHub#!/usr/bin/env python3
"""Publish an approved showcase submission (run by showcase-publish.yml).
流程:解析投稿 issue 的表单正文 → 下载视频(issue 附件或外部直链)→
ffmpeg 规整成 h264 mp4 + 抽 poster → 转存到 showcase-media release →
更新 release 上的 showcase.json → 触发 deploy-pages 重新部署 → 回帖并关闭
issue。投稿媒体与 showcase.json 全程只存 release 资产,不进 git。
字段解析锚点是 issue form 各字段的 label(GitHub 渲染为 "### <label>"),
与 .github/ISSUE_TEMPLATE/showcase.yml 一一对应,两边必须同步改。
环境变量(workflow 注入):GH_TOKEN GH_REPO ISSUE_NUMBER ISSUE_BODY
ISSUE_USER ISSUE_CREATED_AT RUN_ID
"""
import json
import os
import re
import subprocess
import sys
import tempfile
from datetime import datetime, timezone
from pathlib import Path
RELEASE_TAG = 'showcase-media'
MAX_BYTES = 300 * 1024 * 1024 # 单支投稿上限 300MB
SHOWCASE_URL = 'https://vincentwei1021.github.io/video-shotcraft/showcase.html'
# 与 showcase.yml 各字段 label 一致
FIELD_TITLE = '作品名'
FIELD_DESC = '一句话介绍'
FIELD_CARDS = '用到的镜头卡(可选)'
FIELD_FILE = '视频文件(≤10MB 直接拖拽上传)'
FIELD_URL = '视频外部链接(大文件用)'
FIELD_X = 'X 账号(可选)'
FIELD_YOUTUBE = 'YouTube(可选)'
FIELD_XHS = '小红书(可选)'
FIELD_DY = '抖音(可选)'
ATTACHMENT_RE = re.compile(
r'https://github\.com/user-attachments/(?:assets/[\w-]+|files/\d+/\S+?)(?=[)\s"\']|$)'
r'|https://user-images\.githubusercontent\.com/\S+?(?=[)\s"\']|$)')
URL_RE = re.compile(r'https?://\S+?(?=[)\s"\']|$)')
YOUTUBE_RE = re.compile(r'^https?://(?:www\.)?(?:youtube\.com|youtu\.be)/', re.I)
def run(cmd, **kw):
print('+', ' '.join(str(c) for c in cmd), flush=True)
return subprocess.run(cmd, check=True, **kw)
def parse_form(body):
"""Issue form 正文按 '### <label>' 切段,值为段内文本。"""
sections = {}
current = None
for line in body.replace('\r\n', '\n').split('\n'):
if line.startswith('### '):
current = line[4:].strip()
sections[current] = []
elif current is not None:
sections[current].append(line)
fields = {}
for key, lines in sections.items():
value = '\n'.join(lines).strip()
fields[key] = '' if value == '_No response_' else value
return fields
def pick_video_url(fields):
attachment = ATTACHMENT_RE.search(fields.get(FIELD_FILE, ''))
if attachment:
return attachment.group(0)
external = URL_RE.search(fields.get(FIELD_URL, ''))
if external:
return external.group(0)
sys.exit('投稿里既没有视频附件也没有外部链接,无法发布')
def probe_codec(path):
result = subprocess.run(
['ffprobe', '-v', 'error', '-select_streams', 'v:0',
'-show_entries', 'stream=codec_name', '-of', 'csv=p=0', str(path)],
capture_output=True, text=True)
if result.returncode != 0 or not result.stdout.strip():
sys.exit(f'ffprobe 认不出下载的文件,不是可用的视频:{result.stderr.strip()}')
return result.stdout.strip().splitlines()[0]
def normalize_video(raw, out_mp4):
"""h264 直接 remux(无损 + faststart),其余编码统一转码成 1080p 内 h264。"""
codec = probe_codec(raw)
if codec == 'h264':
remux = subprocess.run(
['ffmpeg', '-y', '-i', str(raw), '-c', 'copy',
'-movflags', '+faststart', str(out_mp4)])
if remux.returncode == 0:
return
run(['ffmpeg', '-y', '-i', str(raw),
'-vf', "scale='min(1920,iw)':-2",
'-c:v', 'libx264', '-crf', '20', '-preset', 'veryfast',
'-pix_fmt', 'yuv420p', '-c:a', 'aac', '-b:a', '192k',
'-movflags', '+faststart', str(out_mp4)])
def extract_poster(video, poster):
for seek in ('1', '0'):
result = subprocess.run(
['ffmpeg', '-y', '-ss', seek, '-i', str(video), '-frames:v', '1',
'-vf', "scale='min(960,iw)':-2", str(poster)])
if result.returncode == 0 and poster.exists() and poster.stat().st_size > 0:
return
sys.exit('无法从视频抽取 poster 帧')
def main():
env = os.environ
issue = int(env['ISSUE_NUMBER'])
fields = parse_form(env.get('ISSUE_BODY', ''))
title = fields.get(FIELD_TITLE, '').strip()
description = fields.get(FIELD_DESC, '').strip()
if not title or not description:
sys.exit('投稿缺少作品名或介绍,请补全后再加 approved 标签')
video_url = pick_video_url(fields)
work = Path(tempfile.mkdtemp(prefix='showcase-'))
raw = work / 'raw-video'
# -A 给个浏览器 UA:部分网盘直链对空 UA 返回 403
run(['curl', '-fSL', '--retry', '3', '--max-time', '600',
'--max-filesize', str(MAX_BYTES),
'-A', 'Mozilla/5.0 (video-shotcraft showcase bot)',
'-o', str(raw), video_url])
if raw.stat().st_size > MAX_BYTES:
sys.exit(f'视频超过 {MAX_BYTES // 1024 // 1024}MB 上限')
mp4 = work / f'showcase-{issue}.mp4'
poster = work / f'showcase-{issue}.jpg'
normalize_video(raw, mp4)
extract_poster(mp4, poster)
# release 首次投稿时自动创建
exists = subprocess.run(['gh', 'release', 'view', RELEASE_TAG],
capture_output=True)
if exists.returncode != 0:
run(['gh', 'release', 'create', RELEASE_TAG,
'--title', 'Showcase media assets',
'--notes', 'User-submitted showcase videos + showcase.json '
'(published by showcase-publish.yml, not tracked in git).'])
manifest_path = work / 'showcase.json'
fetched = subprocess.run(
['gh', 'release', 'download', RELEASE_TAG, '--pattern', 'showcase.json',
'--dir', str(work), '--clobber'], capture_output=True)
if fetched.returncode == 0 and manifest_path.exists():
manifest = json.loads(manifest_path.read_text(encoding='utf-8'))
else:
manifest = {'items': []}
now = datetime.now(timezone.utc).isoformat(timespec='seconds')
youtube = fields.get(FIELD_YOUTUBE, '').strip()
xiaohongshu = fields.get(FIELD_XHS, '').strip()
# 兼容 YouTube 字段上线前把频道链接填到“小红书”的投稿。
if not youtube and YOUTUBE_RE.match(xiaohongshu):
youtube, xiaohongshu = xiaohongshu, ''
entry = {
'id': issue,
'title': title,
'description': description,
'cards': fields.get(FIELD_CARDS, '').strip(),
'author': {
'github': env.get('ISSUE_USER', ''),
'x': fields.get(FIELD_X, '').strip(),
'youtube': youtube,
'xiaohongshu': xiaohongshu,
'douyin': fields.get(FIELD_DY, '').strip(),
},
# ?v= 缓存穿透:重新发布同一 issue 时 Pages CDN 才会取新文件
'video': f'./showcase-media/showcase-{issue}.mp4?v={env.get("RUN_ID", "0")}',
'poster': f'./showcase-media/showcase-{issue}.jpg?v={env.get("RUN_ID", "0")}',
'submittedAt': env.get('ISSUE_CREATED_AT', ''),
'approvedAt': now,
}
manifest['items'] = ([entry]
+ [item for item in manifest['items'] if item.get('id') != issue])
manifest['updatedAt'] = now
manifest_path.write_text(
json.dumps(manifest, ensure_ascii=False, indent=2) + '\n', encoding='utf-8')
run(['gh', 'release', 'upload', RELEASE_TAG, str(mp4), str(poster),
str(manifest_path), '--clobber'])
# GITHUB_TOKEN 的 push 不触发 workflow,但显式 workflow_dispatch 可以
run(['gh', 'workflow', 'run', 'deploy-pages.yml'])
run(['gh', 'issue', 'comment', str(issue), '--body',
f'🎉 已发布!你的作品几分钟后会出现在 [Showcase 页]({SHOWCASE_URL})。\n\n'
f'需要修改或下架,随时在本 issue 留言或开新 issue。感谢投稿!'])
run(['gh', 'issue', 'close', str(issue), '--reason', 'completed'])
if __name__ == '__main__':
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |