- Home
- Skills
- Image & Video
- Native Subtitle Quote Image
Native Subtitle Quote Image
Turn a video's real frames into compact 3:4 subtitle quote images, keeping the burned-in subtitles exactly as they are.
$9.99
Works with the AI Agents you already use
Native Subtitle Quote Image
Example session with this skill installed
I have a local video 'interview.mp4'. I want to create a 3:4 subtitle stitch in Native Mode for the segment between 01:20 and 01:45 where the speaker discusses AI safety.
- Read your context and instructions
- Compiled the native subtitle quote
- Generated the image
I have processed the segment from 'interview.mp4' and identified five stable frames with clear hardcoded subtitles. I used Native Subtitle Mode to preserve the original typography and layout.
Subtitle Quote Image: AI Safety Discussion
Source File: interview.mp4
Mode: Native Subtitle (Preserved)
Layout: 3:4 Fixed (1440x1920)
Output Path: ./output/ai_safety_native_stitch.jpg
| Frame | Timestamp | Content Description |
|---|---|---|
| Hero | 00:01:22.4 | Subject centered, looking at camera |
| Strip 1 | 00:01:25.8 | "The core challenge of alignment..." |
| Strip 2 | 00:01:29.2 | "...is ensuring intent matches execution." |
| Strip 3 | 00:01:34.5 | "We cannot leave this to chance." |
| Strip 4 | 00:01:42.1 | "Safety is a feature, not an afterthought." |
The final image has been verified for
- Frame Stability: No motion blur or transition artifacts detected.
- Subtitle Integrity: All text bands are centered and fully legible within the 3:4 crop.
- Visual Consistency: Consistent scaling applied across all subtitle strips.
Next steps
- Review the generated
final_contact_sheet.jpgto verify the narrative flow. - If the crop is too tight on the speaker, I can re-render with
--fit padto add letterboxing. - Run the script with
--render-scriptif you decide to replace these subtitles with a translated version.
ai_safety_native_stitch.jpg
PNG · 1536×1024
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Give your agent a local video or a YouTube link and get social-ready subtitle quote images built from real frames. The skill finds the strongest lines in the transcript, grabs frame-accurate stills with FFmpeg, crops the subtitle strips, and stacks them under a hero shot in a tight 3:4 or original-ratio layout with no gaps. Then it checks every image before handing it over. There are two modes, and they never mix. Native mode keeps the subtitles already burned into the video pixels: no OCR, no redrawing, no rewording. Script mode draws your reviewed lines or translations onto real frames and labels them as added subtitles. The skill asks which mode you want before it starts. It works with what you already have. Local files need only Python, Pillow and FFmpeg. URL mode uses yt-dlp. MIT-licensed, with readable scripts and every network call disclosed.
No DRM/paywall bypass. Requires local CJK fonts for Script mode. Native mode cannot edit/translate burned-in text. Online videos may require user-provided cookies for age/auth restrictions.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago
Requires
Permissions
Local-video mode makes no network calls. URL mode: yt-dlp downloads metadata, subtitles, and video only from links you provide. Runs FFmpeg locally (argument list, no shell) for frame extraction; runs --version on local tools for the read-only environment check. Never reads browser cookie stores. If YouTube requires sign-in, the agent stops and asks for a local video file or a cookies.txt you exported yourself, passed to yt-dlp --cookies for that URL only and never displayed, copied, or uploaded. No environment variables, API keys, or secrets are used; no cache or config files are written. Dependency installs (pip) happen only after the agent explains why and you approve.