Settings

Keys, models and budgets. Nothing here is needed to read the book or do the exercises — only to have Claude explain a mistake or mark your writing.

No database is configured, so settings cannot be saved here. Set ANTHROPIC_API_KEY, ELEVENLABS_API_KEY and APP_SECRET in the environment instead — the app reads those directly.

API keys

Stored encrypted and never sent back to this page — you can replace a key, but not read it. Both providers are only called from the server.

Claudesk-ant-…TAAA · from the environment

ElevenLabssk_5af8…1334 · from the environment

Speech

VoiceAn ElevenLabs voice id, for new speech only.Speaking speedHow fast new speech is spoken. The book's own recordings are already made; use the player's speed control for those.0.90×Keep recordingsOff: a recording is deleted once transcribed.

Budgets

A call that would cross the cap is refused before it is made. Everything graded locally keeps working either way.

Spend is not being counted on this deployment, so these caps are advisory rather than enforced. Counting needs a database and a signed-in account.

Claude, per month$ElevenLabs, per month$

Models

Model ids, not a fixed list — a newer model is a text change here.

Explaining a wrong answerMarking SchreibenMarking SprechenGenerating new itemsChecking generated itemsThe weekly note

Study

Explanations inItems a day