Add persian-tutor: Gradio-based GCSE Persian language learning app

Vocabulary study with FSRS spaced repetition, AI tutoring (Ollama/Claude),
essay marking, idioms browser, Anki export, and dashboard. 918 vocabulary
entries across 39 categories. 41 tests passing.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

python/persian-tutor/CLAUDE.md

@@ -0,0 +1,38 @@
+# Persian Language Tutor
+
+## Overview
+Gradio-based Persian (Farsi) language learning app for English speakers, using GCSE Persian vocabulary (Pearson spec) as seed data.
+
+## Tech Stack
+- **Frontend**: Gradio (browser handles RTL natively)
+- **Spaced repetition**: py-fsrs (same algorithm as Anki)
+- **AI**: Ollama (fast, local) + Claude CLI (smart, subprocess)
+- **STT**: faster-whisper via sttlib from tool-speechtotext
+- **Anki export**: genanki for .apkg generation
+- **Database**: SQLite (file-based, data/progress.db)
+- **Environment**: `whisper-ollama` conda env
+
+## Running
+```bash
+mamba run -n whisper-ollama python app.py
+```
+
+## Testing
+```bash
+mamba run -n whisper-ollama python -m pytest tests/
+```
+
+## Key Paths
+- `data/vocabulary.json` — GCSE vocabulary data
+- `data/progress.db` — SQLite database (auto-created)
+- `app.py` — Gradio entry point
+- `db.py` — Database layer with FSRS integration
+- `ai.py` — Dual AI backend (Ollama + Claude)
+- `stt.py` — Persian speech-to-text wrapper
+- `modules/` — Feature modules (vocab, dashboard, essay, tutor, idioms)
+
+## Architecture
+- Single-process Gradio app with shared SQLite connection
+- FSRS Card objects serialized as JSON in SQLite TEXT columns
+- Timestamps stored as ISO-8601 strings
+- sttlib imported via sys.path from tool-speechtotext project