In Development August 2026

LearnSpanish

A repo-driven Anki deck pipeline for Latin American Spanish — every word gets a picture, native-quality TTS audio, and an example sentence, pushed into Anki without ever touching review history.

pythonankilanguage-learningttspipeline

Screenshots

About

LearnSpanish treats a vocabulary deck like a codebase. Word batches live in YAML files as the single source of truth — word, example sentence, image query, notes — and a pipeline of Python scripts turns each entry into a complete Anki card: a stock photo or generated flat illustration, TTS audio for both the word and the sentence, and the example in context. Everything syncs into Anki via AnkiConnect, keyed on a stable word ID, so edits update cards in place and review scheduling is never lost.

New batches are scaffolded from a frequency-ordered backlog, so the deck grows through the most useful words first. The intended workflow is conversational: open the repo in Claude Code and say “add the next batch.”

Pipeline

  1. new_batch — scaffold the next N words from the frequency list
  2. fetch_images / illustrations — stock photos, or generated flat art for abstract words
  3. gen_audio — TTS for words and sentences
  4. review_images — a generated HTML page to eyeball every picture before it ships
  5. sync — idempotent push into Anki via AnkiConnect

Technical Details

  • Language: Python; YAML word data, HTML review page
  • Anki integration: AnkiConnect add-on; FSRS scheduling on the deck; AnkiWeb sync to mobile
  • Images: Pixabay/Openverse search plus programmatic illustrations