memento-flashcards

Create and review spaced-repetition flashcards with agent-graded answers and YouTube quiz generation.

Updated Jun 5, 2026
One-click install
npx skills add https://github.com/xu1713/openhorse --skill memento-flashcards-xu1713
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memento-flashcards
Source: https://github.com/xu1713/openhorse/tree/main/openhorse/openhorse/optional-skills/productivity/memento-flashcards
Command: npx skills add https://github.com/xu1713/openhorse --skill memento-flashcards-xu1713

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires youtube-transcript-api, and includes scripts (resource) components.

What problem does it solve? Remembering facts long-term requires structured review, but manually building flashcard decks and scheduling reviews is tedious. This Skill provides a local, file-based flashcard system with spaced-repetition scheduling so facts, notes, and video content become reviewable cards managed entirely by the agent. ## Core Features & Use Cases - Flashcard Creation from Facts: Turn any statement into a Q/A card stored in a local JSON file, organized into collections. - Spaced-Repetition Review: Review due cards with free-text answers graded by the agent, with adaptive intervals (hard/good/easy) and automatic retirement after three consecutive easy ratings. - YouTube Quiz Generation: Fetch a video transcript and generate a 5-question quiz stored as flashcards, deduplicated by video ID. - Deck Management: Export and import decks as CSV, view statistics, and delete cards or collections. - Use Case: After watching a lecture on YouTube, ask for a quiz on the video; the Skill fetches the transcript, creates five Q/A cards, and quizzes you one question at a time with graded feedback. ## Quick Start Ask the agent to save a fact as a flashcard or say "review my due cards" to start a spaced-repetition session.

Frequently Asked Questions about memento-flashcards

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create flashcards from facts with spaced repetition?

State a fact and ask the agent to save it as a flashcard; it generates a question-answer pair and stores it via the memento_cards.py add command. Cards are scheduled for review automatically, with intervals of 1, 3, or 7 days based on how you rate your recall.

How to generate a quiz from a YouTube video transcript?

Send a YouTube URL and ask for a quiz. The youtube_quiz.py script fetches the transcript using youtube-transcript-api, the agent generates five Q/A pairs from the first 15,000 characters, and the cards are stored with deduplication by video ID.

Does the YouTube quiz feature require extra dependencies?

Yes, fetching transcripts requires the youtube-transcript-api Python package, installed with pip install youtube-transcript-api. The core flashcard storage and review features use only the Python standard library and need no external packages.

Can I export and import flashcard decks as CSV?

Yes, the export subcommand writes all cards to a three-column CSV (question, answer, collection), and the import subcommand reads the same format. Rows missing a collection column fall back to the collection name passed via the --collection argument.

Why is the YouTube transcript fetch failing?

Failures occur when the video has no English transcript, transcripts are disabled, or youtube-transcript-api is not installed. The script returns a structured error indicating missing_dependency or transcript_unavailable so the cause can be identified.

When should cards not be created automatically?

Cards are only created on explicit request or after confirmation when a fact is shared without mentioning flashcards. General questions, coding tasks, and normal conversation never trigger card creation, keeping the deck free of unintended entries.