mcpField
F
Flash

Flash

Remote
flashmemorize.comScience & Education7Updated 4d ago
RepositoryVisit website

About Flash

Flash is a self-hosted spaced-repetition server that lets AI clients create cards, quiz users, and schedule reviews with FSRS. It is for learners who want to study through an MCP client and manage their decks on their own server.

Key features

  1. 01Create cards through AI clients
  2. 02Quiz users on their decks
  3. 03Schedule reviews with FSRS
  4. 04Import and export Anki .apkg decks
  5. 05Run as one Rust binary with one data folder

Use cases

  1. 01Turn AI-generated study material into deck cards
  2. 02Ask an AI client to quiz a deck
  3. 03Review cards according to an FSRS schedule
  4. 04Move existing Anki decks into or out of Flash

MCP support

Flash provides an MCP endpoint for ChatGPT, Claude, and other MCP clients to create cards and quiz users. The self-hosted server connects through its `/mcp` URL.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://flashmemorize.com/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Flash

Your AI makes the flashcards, Flash makes sure you remember them.

Hosted version · Every setting · MCP tools · Changelog · Contributing

An AI chat adds eight cards to a World capitals deck in Flash, then one card is studied: Show answer, then Good

A self-hosted spaced-repetition server. Connect ChatGPT, Claude or any MCP client: it writes cards into your decks and quizzes you, and Flash schedules every review with FSRS. One Rust binary, one data folder. Imports and exports Anki .apkg decks.

Run it

docker run -d --name flash -p 127.0.0.1:8437:8437 -v flash-data:/data ghcr.io/flash-cards/flash
docker logs flash | grep enroll

Open the enroll link at http://localhost:8437 and you are the admin; invite everyone else from Settings. To serve it beyond your machine, put a TLS reverse proxy in front and set FLASH_BASE_URL to its address (docker-compose.yml is the hardened setup). Your data is the flash-data volume: back it up, and upgrade by pulling a newer image and recreating the container. Locked out? docker exec flash flash-server reset-link [email protected] prints a link to choose a new password (with compose: docker compose exec flash ...).

Installed with docker compose (as up to 0.3.0)? Your data is in that folder's volume (docker volume ls, e.g. flash_flash-data), not in flash-data. Upgrade from the same folder, backing up first, since upgrades only go forward:

docker compose stop flash
docker run --rm -v flash_flash-data:/data -v "$PWD":/backup debian:bookworm-slim \
    tar czf /backup/flash-data.tgz -C /data .
docker compose pull && docker compose up -d

Connect your AI

Add https://<your host>/mcp as a custom connector in ChatGPT, Claude, Grok, Le Chat or Gemini (OAuth, no keys to copy). In a terminal: claude mcp add --transport http flash https://<your host>/mcp. Then say "quiz me on my deck".

The hosted version adds card generation from your files and grading of spoken answers.

License

AGPL-3.0-only, with embedded assets in THIRD_PARTY.md. Contributions are accepted under the CLA.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.