AnkiWeb addon 1135432140
Analyzes recently missed cards from a chosen deck and time window, using an OpenRouter API key to summarize weak areas and suggest card improvements.
Open on AnkiWeb GitHub Find alternatives
active
| Min Anki | Max Anki | Updated |
|---|---|---|
| 2.1.1 | + | 2026-07-04 |
Loading…
Finds CC-CEDICT Chinese words using only characters from selected decks, sorted by frequency, plus missing HSK words, and analyzes character frequency.
Creates focused cram sessions from recently missed or hard cards, using six confidence-based decks, smart spacing, grading controls, progress tracking, deck filters and persistent progress.
Identifies high-yield weak areas from Again/Hard review patterns, recommends linked resources, saves study plans, and generates targeted UWorld or AMBOSS practice-question sets.
Inserts synthetic review log entries for chosen dates, with configurable deck, card text, day start, timing, and spacing, so heatmaps and statistics show restored streaks.
Quickly finds recently saved media files within a time threshold and inserts them into Anki fields, optionally renaming and moving or copying them.
Parses recent Anki reviews and displays your most-missed tags to guide study, with real-time results, configurable settings, danger scores, and AnKing-based UWorld practice blocks.
Anki already shows you what cards you get wrong. AnkiLens helps you find patterns in the cards you keep missing.
Instead of digging through hundreds of reviews, get practical recommendations on what to study next and how to improve your cards. Improving the quality of your Anki sessions will help you learn more with less time.
AnkiLens is especially useful for learners in medicine, law, nursing, and other high-volume subjects where large decks and daily reviews can get overwhelming.
In Anki, go to Tools → Add-ons → Get Add-ons… and paste this code: 1135432140
To use AI features, you’ll need an OpenRouter API key.
AnkiLens uses OpenRouter to generate AI insights. Card content from the selected analysis may be sent to OpenRouter and the selected model provider for processing. Only analyze cards you are comfortable sending to an external AI service.
Coming soon!
AnkiLens is a read-only Anki add-on for reviewing missed cards.
Pick a deck, choose a time window, and AnkiLens shows a short summary of what has been going well and which missed cards may be worth checking next. It uses your recent review history and missed-card text, but it does not change Anki scheduling or edit any cards.
This alpha uses a bring-your-own OpenRouter API key for generated insights.
Symlink the add-on into Anki:
ln -s "$(pwd)/addon" "$HOME/Library/Application Support/Anki2/addons21/ankilens"
Restart Anki after Python changes.
AnkiLens needs an OpenRouter API key for generated insights.
Create a key at:
https://openrouter.ai/keys
Advanced users can also use the OPENROUTER_API_KEY environment variable:
export OPENROUTER_API_KEY="your-key-here"
open -a Anki
On macOS, launching Anki from the Dock may not include shell environment variables. If insights do not load, quit Anki and start it from Terminal with the commands above.
You can also set the key in Anki:
Tools -> AnkiLens -> Set API key
Paste the key into the dialog and AnkiLens will save it locally.
The default model is deepseek/deepseek-v4-flash. Advanced users can change
llm_model in addon/config.json, but the main alpha does not include an
in-app model picker.
OpenRouter usage and billing stay in the user's OpenRouter account. AnkiLens does not show token pricing in the UI.
When generated insights are enabled, AnkiLens sends selected missed-card text and review metadata for the chosen deck/window to the configured LLM endpoint. It does not send the whole collection.
The add-on is read-only. It does not change scheduling, notes, cards, decks, or review logs.
Configuration lives in addon/config.json.
Useful settings:
llm_summary_enabled: turns generated insights on or off.llm_model: model name.llm_api_url: chat-completions compatible API endpoint.llm_api_key_env: environment variable that contains the API key.llm_api_key: local API key. Do not commit a real value.llm_max_cards: maximum missed cards sent for summarization.llm_max_chars: maximum prompt size.demo_data_enabled: adds local demo review data.count_hard_as_miss: treats both Again and Hard as misses by default. Set to false for Again-only.The default API key environment variable is:
OPENROUTER_API_KEY
Do not commit API keys.
Run tests with:
make test
The tests do not require a live Anki GUI.
To check LLM output against realistic missed-card content without opening Anki:
OPENROUTER_API_KEY="your-key-here" make llm-check
You can also pass your own card sample:
OPENROUTER_API_KEY="your-key-here" python3 scripts/check_llm_output.py --input cards.json
The JSON file should be a list of cards:
[
{
"label": "Smooth ER role",
"text": "Front: Smooth ER role. Back: lipid synthesis, detoxification, calcium storage.",
"misses": 3,
"reviews": 5
}
]
Build an uploadable add-on package with:
make package
The package is written to dist/ankilens.ankiaddon.
addon/anki_entry.py - Anki menu, toolbar, page loading, and webview messages.addon/anki_gateway.py - reads Anki review logs and cleans card text.addon/anki_browser.py - builds Browse searches and opens Anki Browse.addon/analytics.py - turns review logs into missed-card summaries.addon/debrief.py - small data models for evidence and generated insights.addon/debrief_page.py - the Insights page HTML/CSS/JS.addon/llm_summary.py - prompt, API request, and response parsing.addon/demo_data.py - local demo review history.scripts/check_llm_output.py - queries the configured LLM with sample card content.tests/ - unit tests for the core behavior.docs/ - notes for verification and future work.