Why local-first matters
Many paraphrase services ask writers to paste drafts into a remote form and pay for stronger limits. Lugas takes a different route. The public app is static, installable, and free. It can build a prompt without any endpoint, or call only the endpoint you enter.
Output can be Indonesian (EYD V) or English, independent of the interface language. The meaning check runs in the browser without AI: numbers with decimal commas, statistics, citations, quotations, negation in Indonesian and English, hedges, association versus causation, direction, and scope. The app stores no manuscripts on a Lugas server because the web version has no manuscript server.
Safe setup choices
| Choice | How it works | Privacy check |
|---|---|---|
| Prompt only | Lugas builds a compact prompt. Paste the model's answer into the rewrite box and press Check rewrite. | No AI call leaves the browser. |
| Local endpoint | Use an OpenAI-compatible local server such as a local model gateway. | Keep endpoint on your own machine or network. |
| Provider endpoint | Enter your own endpoint, model, and token when your provider permits browser calls. | Your text goes to that provider, not to Lugas. |
Why there is no daily cap
Endpoint calls use your own model, token, and quota, so a browser-side cap protects nobody's costs. Earlier versions limited calls to three per day; that cap was removed. Requests stop after 90 seconds and can be cancelled.
A true shared-key hosted service needs a backend. The planned design is a small Worker API, a D1 table for anonymous daily quota, a KV cache for rate-limit decisions, Turnstile for abuse resistance, provider tokens stored only as server secrets, and manuscript logging disabled by default. That backend is intentionally separate from the current static app.