- Home
- Skills
- Education & Learning
- Scene Audio Flashcards — Learn with Real Conversations
More screenshots
Works with the AI Agents you already use
Scene Audio Flashcards — Learn with Real Conversations
Turn favorite film, TV and podcast exchanges into multilingual audio flashcards for listening, shadowing and retelling, with original voices and concise cues.
Free
Scene Audio Flashcards — Learn with Real Conversations
Example session with this skill installed
Using the supplied B for Busy recording, make three Shanghainese audio flashcards with English explanations: two everyday exchanges and one sharp, witty reply. Keep enough dialogue to understand the intent. Put concise scene and expression cues on the back, with the full script revealed only when needed.
- Read your context and instructions
- Compiled the scene audio flashcards
- Generated the document
Three-card sample: B for Busy
Target language: Shanghainese. Explanation language: English.
This is the content handoff from an existing three-card demo. The original-audio player is available at https://luna-j.github.io/travel-flashcards/demos/shanghainese.html . The installable skill itself contains no third-party film audio.
| Card | Situation | Practice goal |
|---|---|---|
| Checking before helping | Bai agrees to help on a film set and checks what the job involves. | Check a detail, then confirm you can help. |
| Finding out about someone | Ms Li discovers Bai used to have a job; a small tease keeps the conversation going. | Show surprise and ask how someone knows about you. |
| A sharp answer, a warm comeback | A self-deprecating joke about life going downhill becomes a playful invitation. | Catch the double meaning and respond warmly. |
Practice flow
Listen to the original clip on the front. Flip for the scene, purpose and two short expression cues. Open the dialogue, translation and notes on demand. For example, the first card uses 对吗? to check understanding and 可以啊 to agree.
Delivered files
index.html, library.html, cards.json, audio/, content.md, sources.md and report.md. Optional Anki export includes notes.tsv, flat media and note templates.
Verification
The three clips passed decoding. The on-screen text is used as a meaning cue; dialect transcription, exact clip boundaries and human listening remain unverified. These labels are retained in the source report. Try the samples and confirm length and difficulty before requesting a full batch.
shanghainese.html
PDF · document
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
A conversation you enjoy can become a small language practice session. Scene Audio Flashcards guides your AI agent to select short, useful exchanges from your media, prepare a few playable samples, and agree on difficulty and length before producing the full set.
Listen to the original voice, flip for the situation and a few expression cues, then try saying it yourself. Open the full dialogue, translation and notes when you want more support. Keep practice small enough to return to, with time left to listen and speak.
Use a local recording, podcast RSS feed, episode page or accessible video link. Target and explanation languages are independent, including combinations such as Japanese through English or Shanghainese through English.
Features
- Curates exchanges for the learner's listening comfort, interests and practical speaking goals.
- Prepares two or three playable samples before an approved batch.
- Uses original audio by default, with exact recording ranges and explicit verification states.
- Generates a responsive HTML deck with sound-first cards, concise contextual cues and on-demand dialogue.
- Includes shuffle, theme switching, local progress and progress export/import in the existing player.
- Produces editable JSON and structured Markdown study notes, source information and a QA report.
- Optionally exports Anki TSV with flat media and note templates.
Prerequisites
An AI agent with file writing and local shell execution, Python 3.10+, FFmpeg for audio clipping, a browser and usable recordings. For Agensi delivery, the agent must first restore all supplied text assets into a local folder, preserving scripts and assets paths. Isolated single-file execution alone does not support this package's HTML template lookup and sibling-module imports.
Acquisition, subtitle access, OCR and ASR depend on available host tools. Online sources may require network access. Python helpers and the HTML player do not require API keys, npm or a backend. Optional services and the agent itself may incur costs.
Included files
SKILL.md at the ZIP root, English and Chinese READMEs, MIT license, Python scripts, HTML template, references, an authored text-only example, package allowlist and tests.
Media availability depends on the source and environment. No third-party film or podcast audio is included. Transcript quality, cut boundaries and human listening are reported separately. The Anki export uses TSV and media, and .apkg generation is not included. Personal progress is browser-local, with manual export/import.
The package is prepared for local folder restoration and validation. Agensi's authenticated delivery and review have not yet been exercised. Preserve the existing MIT license and check the platform's current submission terms when listing the open-source package.
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
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 5 days ago
- Free with an Agensi account
Requires
Permissions
Reads supplied recordings and transcripts, writes selected clips and study files, and uses a local shell for Python and FFmpeg. Network is needed for online sources; local material can be processed offline. No mandatory third-party account.