Produce a structured listening comprehension drill from data/listening.json, including transcript segments, key vocabulary callouts, comprehension questions, and a summary task.
Invoke this skill when a learner runs /jp-listen or when generate-daily-tasks assigns a listening task. It structures a listening item into an interactive drill format with clear stages.
| Parameter | Type | Required | Description |
|---|---|---|---|
level | string | yes | Learner's JLPT level: "N5", "N4", "N3" |
topic | string | no | Topic filter (e.g., "greetings", "travel"). Omit for any topic. |
duration_target | integer | no | Max drill duration in seconds. Used to filter items by duration_seconds. |
item_id | string | no | Specific listening item ID to use. If omitted, one is selected by level and topic. |
A structured drill object:
| Field | Type | Description |
|---|---|---|
item_id | string | The listening item used |
title | string | Title of the listening passage |
pre_vocab | object[] | 3–5 key vocabulary items to preview before listening |
transcript_segments | string[] | Transcript split into logical chunks for staged listening |
comprehension_questions | object[] | Questions from the item's questions field |
summary_task | string | Free-production task: learner summarizes the passage in Japanese |
full_transcript_jp | string | Full transcript revealed after drill completion |
full_transcript_zh | string | Full Chinese translation for self-check |
If item_id is not provided:
data/listening.json by level and topic (if given).duration_seconds <= duration_target (if given).storage/daily_state.json completed tasks).Stage 1 — Preview vocabulary: Show pre_vocab items (3–5 key words from key_vocab). Give kana, meaning_zh, and one example sentence.
Stage 2 — First listen: Present the transcript in 2–3 segments. Instruct the learner to read (or listen) without translation and note what they understand.
Stage 3 — Comprehension questions: Ask all questions from questions[]. Learner answers in Japanese. No peeking at transcript yet.
Stage 4 — Second listen with review: Reveal the transcript segments again. Learner compares their understanding to the text.
Stage 5 — Summary task: Learner writes or speaks a 2–3 sentence summary in Japanese. This summary can be fed to review-answer for correction.
For dialogue transcripts, split by speaker turn. For monologue transcripts, split by sentence groups of 2–3 sentences.
Input:
{ "level": "N5", "topic": "greetings", "duration_target": 60 }
Output (abbreviated):
{
"item_id": "listen_001",
"title": "自己紹介 — Self-Introduction at Work",
"pre_vocab": [
{ "kana": "はじめまして", "meaning_zh": "初次見面", "example_jp": "はじめまして、よろしくお願いします。" }
],
"transcript_segments": [
"はじめまして。わたしはリンと申します。台湾から来ました。",
"今日からよろしくお願いします。日本語はまだ勉強中ですが、頑張ります。"
],
"comprehension_questions": [
{ "q": "この人はどこから来ましたか？", "a": "台湾から来ました。" }
],
"summary_task": "このスピーチについて、日本語で2〜3文でまとめてください。"
}