Loading page...Загружаем страницу...正在加载页面...
AtlasRepo

Sloth Inbox: local AI that gives you fewer things to do

Описание и применение

Локальный AI‑ассистент Sloth Inbox обрабатывает текстовые файлы, создаёт краткие резюме, предлагает категории и извлекает явные действия, уменьшая количество открываемых документов.

Задача

Digital clutter from receipts, meeting notes, and emails creates a tedious decision loop about what matters, where it goes, and whether it requires action.

Результат

In a measured run, six fictional documents were processed in 15.1 seconds, producing concise summaries and only two actionable follow‑ups, with no tasks generated from informational documents.

Рабочий процесс

  • Drop notes, receipts, or emails into the inbox folder
  • Ollama reads the files using Gemma 3 4B
  • The UI shows a summary, proposed category, and any extracted follow‑ups
  • User reviews the proposal and corrects category if needed
  • User files a named copy, keeping the original untouched
  • User reads the brief instead of reopening the full file

Доказательства

In a measured local run: six fictional documents, processed by gemma3:4b through Ollama in 15.1 seconds.

Источник

The workflow is deliberately small: 1) Drop in notes, 2) Let Sloth read them through Ollama, 3) Glance at a proposal beside the original; correct its category if needed, 4) File a named copy, keeping the original untouched, 5) Read one brief instead of opening the whole pile again.

Источник

The core is Gemma 3 4B running locally through Ollama. The model performs meaning-based classification, summary writing, and request extraction.

Источник

  • Дата записи: 2026-10-24T02:11:13.115Z