- Home
- Skills
- Education & Learning
- English Lesson Video
More screenshots
Works with the AI Agents you already use
English Lesson Video
Turn interviews and podcasts into 9:16 English lesson videos with native-language narration, subtitles, cover art and PDF study notes.
$29
English Lesson Video
Example session with this skill installed
Create a 9:16 English lesson spec for the phrase 'I stopped apologizing' from a podcast, with Chinese narration and a PDF handout.
- Read your context and instructions
- Compiled the english lesson video
I have prepared the SPEC file and the corresponding production workflow for the "I stopped apologizing" lesson.
English Lesson Assets
| File | Status | Description |
|---|---|---|
01_x.py | Generated | Python SPEC for engine rendering |
01_x_handout.md | Generated | Source markdown for PDF generation |
01_x_social.md | Generated | Posting copy for WeChat/Red |
# 01_x.py - Lesson Specification
NAME = '01_x'
SPEC = {
'lang': 'zh',
'cards': [
dict(n='01', title='Core Expression',
lines=[(950, 'I stopped apologizing', 56, 'fg', True),
(1040, 'for wanting more.', 68, 'green', True),
(1220, '我不再为「想要更多」道歉。', 45, 'fg', False)],
speech=[('zh', '注意短语 apologize for。'),
('en', 'apologize for wanting more'),
('zh', '意思是为自己想要更多而感到抱歉。')],
source=(12.40, 15.90)),
],
}
Next steps
- Run
python3 engine.py 01_xto render the MP4 and synthesize narration. - Generate the study guide using
python3 handout.py 01_x_handout.md ./output --lang zh. - Verify the video doesn't exceed
y=1500for platform UI compatibility.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual video editing for language learning content is tedious, requiring precise timing, bilingual overlays, and synchronized narration. This skill automates the production of high-quality, vertical (9:16) English lesson videos derived from long-form interviews or podcasts. It handles the entire pipeline from transcription and segment slicing to rendering narrated cards and generating companion PDF study notes.
What it does
- Segmented Slicing identifies specific English expressions within long audio/video files and isolates them for teaching.
- Automated Card Rendering creates visual slides with configurable text positioning, bilingual styling, and distinct font treatments for English and native translations.
- Multilingual Narration synthesizes localized explanations using ElevenLabs, supporting native languages like Chinese, Japanese, Korean, Spanish, French, German, and Portuguese.
- Study Material Generation produces matching SRT subtitles, social media post copy, and detailed PDF handouts with sentence-by-sentence comparisons.
- Audio Synchronization manages "Slow English" playback for teaching segments followed by full-speed original audio for natural listening practice.
How it works
- Transcription runs the source video through
transcribe.pyusingfaster-whisperto generate time-coded text. - Segmentation uses
split_srt.pyto carve the content into instructional chapters without breaking sentence integrity. - Spec Creation defines the lesson logic in a Python SPEC file, detailing card text, narration scripts, and original audio timestamps.
- Rendering executes
engine.pyto combine visuals, synthesized speech, and source audio into a final 1080x1920 MP4.
Frameworks & tools
The engine relies on FFmpeg for media processing, Pillow for image generation, and faster-whisper for ASR. It integrates with the ElevenLabs API for high-quality TTS and uses a headless Google Chrome instance via handout.py to render PDF study notes.
Why this beats prompting it yourself
Generic AI video generators often fail at the precise synchronization required for "listen-repeat" language pedagogy. This skill enforces specific educational constraints, such as ensuring original audio includes 0.4s trailing buffers and maintaining strict separation between instructional narration and native speaker clips.
Use cases
- Creating English shadowing clips for TikTok, Reels, or YouTube Shorts.
- Converting podcast episodes into structured daily vocabulary lessons.
- Generating bilingual study handouts for classroom or remote teaching.
- Building a library of "micro-lessons" from celebrity interview archives.
Known limitations
Narration synthesis requires a paid ElevenLabs API key. PDF generation requires Google Chrome to be installed, and font paths are currently optimized for macOS system defaults.
Worked example
Input — a 14-minute English podcast interview (input.mp4) with NATIVE_LANG=zh.
What comes out — one 1080x1920 lesson video, one 1080x1440 cover, one study-notes PDF and one SRT. Each chapter interleaves narrated explanation cards with the untouched source clip and bilingual subtitles, so learners hear the original voice instead of TTS-only audio.
Result — a 17-chapter series was produced with this exact code, each chapter 2-15 minutes long. The two screenshots on this page are real frames from chapter 04 (Chinese native language), plus its 8-page PDF.
Run it — clone the repo, cp config.example.env config.env, fill in the key and paths, then follow references/workflow.md (10 steps from transcription to delivery).
Prerequisites
- Python 3.9+ with
requestsandPillow(addfaster-whisperonly if you want automatic transcription) ffmpegandffprobeon PATH- Google Chrome installed (headless PDF export)
- An ElevenLabs API key for narration, or the Agensi voice tool
- Source audio or video you have the rights to use
Included files
SKILL.md— the 10-step workflow and content rulesscripts/engine.py— card renderer, narration cache, ffmpeg assemblyscripts/handout.py— HTML and PDF study notesscripts/covers.py— 1080x1440 cover generatorscripts/transcribe.py,make_srt.py,split_srt.py,fix_srt.py— subtitle pipelinereferences/workflow.md,spec-format.md,languages.mdassets/spec-template.py— copy this to start a new chapterconfig.example.env— every key and path in one place
Requires ElevenLabs API for narration. PDF engine requires local Chrome. Optimized for macOS font paths. Does not package source video for distribution due to copyright constraints.
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 today
Requires
Permissions
Allowed Hosts