thiswillbeyourgithub/AnkiAIUtils

What it solves

This project provides a suite of AI tools designed to improve the quality and effectiveness of Anki flashcards. It specifically targets cards that a user struggles with, automatically enhancing them with explanations, visual aids, and memory hooks to reduce the effort required for manual card refinement during intense study periods.

How it works

The toolkit consists of several Python scripts that modify Anki notes directly in-place. It uses LiteLLM to remain provider-agnostic, allowing users to connect to various LLM and image generation providers (like OpenAI, Anthropic, or Stable Diffusion).

Key mechanisms include:

• Semantic Similarity: Matches cards with relevant examples from user-provided training datasets to ensure consistent AI output.
• Mnemonic Systems: Implements proven techniques like the Major System (converting numbers to sounds/words) and personal memory anchors.
• Direct Integration: Modifies Anki notes across all platforms (Windows, Mac, Linux, Android, iOS) and can be automated via cron jobs.

Who it’s for

It is primarily designed for power users of Anki, particularly students in demanding fields like medicine, who manage large decks and need a scalable way to improve poorly worded cards or create complex memory aids.

Highlights

• Illustrator: Generates custom mnemonic images using DALL-E or Stable Diffusion based on card content.
• Reformulator: Rephrases unclear or poorly worded cards while preserving cloze deletions and media.
• Mnemonics Creator: Generates multiple mnemonic options using the Major System and personal anchors.
• Explainer: Breaks down complex concepts and provides context for cards the user failed.
• Mnemonics Helper: An interactive CLI tool for brainstorming and saving new mnemonics to a dataset.