video-analyze
Video analysis and inspection using FFprobe and FFmpeg. Reads metadata, codec info, resolution, bitrate, duration, HDR status,…
AI video generation from text or image prompts using Google Veo 3.1 Fast ($0.15/sec, native audio, 8s clips, up to 4K), Runway Gen-4 Turbo ($0.05/sec, silent B-roll, 10s), or local Stable Video Diffusion (free, 2-4s). Generates transition clips, B-roll footage, animated
$ npx -y skills add AgriciDaniel/claude-video --skill video-generate --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/video-generateContext preview
The summary Claude sees to decide when to auto-load this skill.
AI video generation from text or image prompts using Google Veo 3.1 Fast ($0.15/sec, native audio, 8s clips, up to 4K), Runway Gen-4 Turbo ($0.05/sec, silent B-roll, 10s), or local Stable Video Diffusion (free, 2-4s). Generates transition clips, B-roll footage, animated
name: claude-video-generate description: > AI video generation from text or image prompts using Google Veo 3.1 Fast ($0.15/sec, native audio, 8s clips, up to 4K), Runway Gen-4 Turbo ($0.05/sec, silent B-roll, 10s), or local Stable Video Diffusion (free, 2-4s). Generates transition clips, B-roll footage, animated intros/outros, establishing shots, and image-to-video animations. Use when user says "generate video", "AI video", "video from text", "video from image", "B-roll", "veo", "runway", "text to video", "image to video", or "create footage". allowed-tools: - Bash - Read - Write
> **Tip:** For advanced prompt engineering, Creative Director workflow, image-to-video with > prompt crafting, domain mode selection, and cost tracking, use the standalone `/veo` skill. > This sub-skill is a lightweight wrapper for quick generation within video pipelines.
1. Check API keys:
2. If no API keys available, only local SVD (very limited) is available 3. Always show estimated cost and confirm before API generation
Generates up to 8-second clips with synchronized audio at up to 4K resolution.
source ~/.video-skill/bin/activate python3 scripts/video_generate.py \ --provider veo \ --prompt "A drone view of the Grand Canyon at sunset with ambient wind sounds" \ --aspect 16:9 --resolution 1080p \ --output broll_canyon.mp4
**Options:**
**Pricing:**
**Features:**
**Generation time:** 2-5 minutes (Fast), 5-12 minutes (Standard). Script polls with progress.
Fastest and cheapest for silent B-roll footage.
source ~/.video-skill/bin/activate python3 scripts/video_generate.py \ --provider runway \ --prompt "Smooth camera pan across a modern office workspace" \ --duration 10 \ --output broll_office.mp4
**Options:**
**Pricing:** $0.05/sec x 10s = **$0.50 per clip**
**Limitations:**
**Best for:** Placeholder B-roll, transitions, background footage where audio isn't needed.
Generates 2-4 second clips from a source image. No text-to-video.
source ~/.video-skill/bin/activate python3 scripts/video_generate.py \ --provider local \ --image source_frame.png \ --output motion.mp4
**Limitations:**
**Best for:** Subtle motion effects from still images when no API keys available.
Extend an existing generated clip:
python3 scripts/video_generate.py \ --provider veo \ --extend existing_clip.mp4 \ --prompt "Continue the camera movement, revealing the cityscape below" \ --output extended_clip.mp4
Can chain up to 20 extensions (8s each = ~148s total). Each extension costs $1.20 (Fast).
1. Veo 3.1 Fast → Best quality, native audio ($1.20/clip) 2. Runway Gen-4 → Budget silent B-roll ($0.50/clip) 3. Local SVD-XT → Free, image-to-video only, 2-4s, degraded quality 4. Remotion → Programmatic motion graphics (existing create sub-skill)
Before ANY API generation, display and confirm:
Provider: Google Veo 3.1 Fast Duration: 8 seconds Resolution: 1080p Estimated cost: $1.20 Proceed? (y/n)
Never generate without explicit user confirmation for paid API calls.
All generated videos are MP4 with H.264 codec. The script runs ffprobe on the output and returns JSON metadata including resolution, duration, codec, and file size.
1. Always confirm API costs before generation 2. Run `bash scripts/preflight.sh` for output path validation 3. Track cumulative API costs across multiple generations in a session 4. Warn if total session cost exceeds $10 5. For batch generation, show total estimated cost upfront
Load `references/video-generation.md` for API setup, polling patterns, cost formulas, and emerging models.
AI-powered video production suite for Claude Code. Edit, transcode, caption, analyze, generate, and create promo videos: all from your terminal.
Video analysis and inspection using FFprobe and FFmpeg. Reads metadata, codec info, resolution, bitrate, duration, HDR status,…
Audio processing for video files using FFmpeg. Loudness normalization to streaming standards (LUFS), noise reduction, audio…
Speech-to-text transcription and animated subtitle generation using Whisper and FFmpeg. Creates word-by-word animated captions…
Programmatic video creation using Remotion (React-based). Generates title cards, data visualizations, motion graphics, branded…
Video downloading via yt-dlp with format selection, quality control, subtitle download, audio-only extraction, and playlist…
Video editing operations using FFmpeg: trimming, cutting, splitting, concatenation, speed changes, cropping, scaling, rotation,…