Add key vocabulary

Sentence cards are easier to study when they highlight a few words that carry the sentence’s meaning or nuance. This recipe fills a Key Vocabulary field with one to three structured HTML entries per note.

Assume the notes have Japanese, English, and Key Vocabulary fields. Add the target field to the note type in Anki before starting.

Export the sentence notes

Open Anki and export the deck:

anki-llm export "Japanese Sentences" --output sentences.yaml

For a narrower job, export an Anki query instead. The export reference covers deck, query, and note type selection.

Start from the full prompt

Download or copy the runnable Key Vocabulary prompt and save it as prompt-key-vocab.md.

Its frontmatter declares the field and requires a tagged result:

---
output:
  field: "Key Vocabulary"
  require_result_tag: true
---

The body asks the model to:

• select one to three useful words for an intermediate learner;
• use dictionary forms in headings and include readings;
• give part of speech, a concise meaning, and sentence-specific context;
• return semantic <h3> and <dl class="vocab-entry"> HTML;
• put only the finished HTML inside <result> tags.

Adapt the final placeholders if your note uses different source field names. See Write prompts for placeholder and result-tag rules.

Preview representative notes

Check the filled prompt first:

anki-llm process-file sentences.yaml \
  --output sentences-key-vocab.yaml \
  --prompt prompt-key-vocab.md \
  --dry-run

Then inspect real model output:

anki-llm process-file sentences.yaml \
  --output sentences-key-vocab.yaml \
  --prompt prompt-key-vocab.md \
  --model gemini-2.5-flash-lite \
  --preview

Look for vocabulary at the intended learner level, correct dictionary forms and readings, valid HTML, and explanations tied to the actual sentence rather than generic dictionary text.

Tip

Preview a varied sample. Include cards with conjugated verbs, kana-only words, idioms, proper names, and sentences where the English translation is loose.

Fill the field

Process the complete export:

anki-llm process-file sentences.yaml \
  --output sentences-key-vocab.yaml \
  --prompt prompt-key-vocab.md \
  --model gemini-2.5-flash-lite

A processed field resembles:

Key Vocabulary: |
  <h3>控える (ひかえる)</h3>
  <dl class="vocab-entry">
    <dt>Type</dt>
    <dd>Ichidan verb</dd>
    <dt>Meaning</dt>
    <dd>To refrain; to hold back</dd>
    <dt>Context</dt>
    <dd>Appears as 控えていて, expressing an ongoing act of restraint.</dd>
  </dl>

The output file retains every original field and saves completed rows incrementally. See Process a file for resume and force behavior.

Review and import

Inspect the HTML in the output file and render several notes if your card CSS has special rules for headings or definition lists. Then open Anki and import:

anki-llm import sentences-key-vocab.yaml --deck "Japanese Sentences"

Caution

The importer writes valid destination fields from the file. Keep the original export and review the processed file before importing a large deck.

If you prefer direct updates, the same prompt works with process-deck. Start with a narrow query and model-backed preview, especially when the target field already contains hand-written notes.