Skip to content

sjquant/cram

v0.1.0MIT

Turn any document into an interactive flashcard quiz in a single self-contained HTML file. Extracts normalized cards — basic, multiple choice and cloze — then renders an offline quiz that remembers your progress and lets you retry only what you missed.

GitHubOpen Website

Cram

Turn any document into a self-graded flashcard quiz — no server, no accounts, just one HTML file.

Cram turns pasted, attached, local, or web material into a validated deck and a self-contained, offline flashcard quiz in a single HTML file you can open in any browser.

Cram demo: run the cram skill, generate validated questions, and study the deck

The generated player supports:

  • basic, multiple-choice (MCQ), and cloze cards;
  • self-grading with a score and review screen;
  • retrying only the cards you missed;
  • optional hints and explanations; and
  • progress persisted in the browser, so you can come back to a deck later.

Try the demo

Open the live demo → — no install, nothing to clone, it just runs in your browser.

The repository also includes the deck that generated it: browse the deck source, or open the rendered HTML file directly from a local clone.

Install in Claude Code

Add the Cram marketplace, then install its cram plugin:

/plugin marketplace add sjquant/cram
/plugin install cram@cram

The first command registers the GitHub marketplace (sjquant/cram). The second follows Claude Code's plugin-name@marketplace-name convention; both names are cram here. After installation, invoke the skill as /cram:cram or ask Claude Code to make flashcards or an interactive quiz from your material.

For the complete extraction, deck-format, and rendering workflow, see the skill guide.

Use with other AI tools

The skill payload follows the portable Agent Plugins format and lives in skills/cram/. The npx skills installer can place it in any supported agent's skill directory (using a symlink by default, or copies when requested):

npx skills add sjquant/cram --skill cram

To target agents explicitly, repeat --agent, for example:

npx skills add sjquant/cram --skill cram \
  --agent codex \
  --agent cursor \
  --agent github-copilot \
  --agent grok \
  --agent kiro-cli \
  --agent antigravity-cli

This repository also includes native marketplace metadata where the format is documented: Claude Code (.claude-plugin/), Codex (.codex-plugin/ and .agents/plugins/), Cursor (.cursor-plugin/), and GitHub Copilot (.github/plugin/). Codex and Copilot can be installed from their respective marketplace catalogs:

codex plugin marketplace add sjquant/cram
codex plugin add cram@cram

copilot plugin marketplace add sjquant/cram
copilot plugin install cram@cram

Kiro Powers and Cursor's public marketplace require their own import or review/publish flow; Antigravity and Grok can use the portable skill, and Grok also reads Claude-compatible plugin marketplaces.

Requirements

For skill users, system python3 is enough—there is nothing to install with pip, and no Node.js, Playwright, Chromium, server, or network connection is needed to play a rendered deck. The output is a single HTML file containing the player and deck data.

The optional Playwright browser suite is for repository contributors and CI only. It requires the JavaScript development dependencies and Chromium described in tests/README.md; skill users never need those dependencies.

License

MIT