Min Anki	Max Anki	Updated
2.1.1	+	2026-05-22

Anki Reflect

A narrative-driven Anki add-on that generates rich, personalized study session reflections — with FSRS metrics, retention trends, and animated growth graphs.

Anki Reflect hooks directly into your local database (mw.col.db) to analyze your daily or weekly study sessions, generating conversational insights and high-quality charts completely offline with zero binary dependencies.

✨ Features

Feature	Details
📊 Session Overview	Cards reviewed, time spent, avg speed, comparison to previous period.
🎯 Retention Analysis	Mature card retention %, correct/total breakdown, and a visual HTML/SVG ring chart.
🔁 Lapse Tracking	"Again" press rate with contextual narrative commentary.
🧠 FSRS Metrics	Median stability, mean difficulty, Estimated Total Knowledge (expected recall count), and struggling card %
📈 Growth Graphs	Pure Python SVG graphs (Mature Card Growth, Daily Review Volume, FSRS Stability Distribution). Fully responsive and crisp on Retina/High-DPI screens without requiring `matplotlib` or compiling C-extensions.
💡 Narrative Insights	Offline, deterministic Python template engine generating conversational, personalized reflections.
🤖 Optional LLM	Plug in an OpenAI or Anthropic key in the config to receive AI-generated narrative reflections.
⬇️ Export	Save your study reflections as a standalone HTML file.

🚀 Installation & Setup

Development Setup (from source)

If you've cloned this repository directly into your Anki add-ons folder, or have symlinked it:

# Symlink this folder into Anki's add-ons directory
ln -s "$(pwd)" ~/Library/Application\ Support/Anki2/addons21/anki_reflect
# (Linux: ~/.local/share/Anki2/addons21/anki_reflect)
# (Windows: %APPDATA%\Anki2\addons21\anki_reflect)

Restart Anki to load the add-on.

🎮 Usage

Once loaded, go to Tools → Generate Session Reflection in your Anki menu.

The report will display in a beautiful, dark-mode glassmorphic interface:

Session Overview — key stats at a glance
Retention & Lapses — accuracy and lapse rate commentary
Metric Breakdown — detailed comparison table of current vs previous sessions
FSRS Insights — deep dive into stability distributions and knowledge accumulation
Mature Card Growth & Volume — clean, responsive timeline charts
Reflection & Insights — final motivational commentary

⚙️ Configuration

Settings can be changed by going to Tools → Add-ons → Anki Reflect → Config:

{
  "report_scope": "today",
  "mature_threshold_days": 21,
  "graph_theme": "dark",
  "llm_enabled": false,
  "llm_provider": "openai",
  "llm_api_key": "",
  "llm_model": "gpt-4o-mini"
}

report_scope: "today", "week", or "month".
mature_threshold_days: The interval threshold (in days) at which a card is classified as mature.
graph_theme: Choose "dark" or "light" to match your system aesthetics.
llm_enabled: Enable to generate narrative paragraphs via OpenAI/Anthropic. Requires a valid API key.

See config.md for more details.

🏗️ Structure

.
├── __init__.py          # Entry point — registers menu action & hooks
├── manifest.json        # Anki add-on metadata
├── config.json          # Default user config
├── config.md            # Config documentation (rendered inside Anki)
├── _version_info.py     # Version specification
├── data/
│   ├── queries.py       # SQL queries targeting mw.col.db
│   └── fsrs_metrics.py  # FSRS parsing and stats calculation
├── report/
│   ├── builder.py       # Report orchestration pipeline
│   ├── graphs.py        # Offline Python SVG graph generator
│   ├── insights.py      # Conditional narrative engine & LLM api client
│   └── templates/
│       ├── report.html.j2 # Jinja2 HTML report template
│       └── styles.css    # Premium glassmorphic styles
└── ui/
    ├── menu.py          # Action hooks
    └── dialog.py        # PyQt WebEngine window

Anki Reflect – Session & Weekly Reflection

AnkiWeb

Maintenance

GitHub

Will it work on my Anki?

Similar addons

README