ahpxex/read-aware

An ebook reader with a self-evolving agent: it remembers your reading across books, and plugins extend the reader and the agent alike.

What it solves

ReadAware addresses the friction of traditional e-readers by integrating an AI agent directly into the reading experience. Instead of treating AI as a sidebar addition, it creates a system where an agent understands the user's books, annotations, and recurring ideas, helping readers dive deeper into unfamiliar passages or manage vocabulary without leaving the page.

How it works

The application uses a local-first architecture where a single agent orchestrates retrieval, context assembly, and memory updates. It utilizes SQLite FTS (Full Text Search) combined with scope, recency, and importance signals for retrieval rather than a vector database. A sandboxed plugin system allows the reader and the agent to be extended with new tools, such as read-aloud voices, dictionaries, and RSS feeds, using a shared domain API layer.

Who it’s for

It is designed for readers across macOS, Windows, Linux, Android, and iOS who want a focused, ADHD-friendly reading environment that supports a wide variety of formats (EPUB, PDF, MOBI, etc.) and an AI-powered assistant to enhance their comprehension and knowledge retention.

Highlights

  • Agent-First Design: An integrated agent that builds an evolving memory of the user's reading habits and notes.
  • Extensible Plugin System: Sandboxed plugins that provide new functionality to both the user and the AI agent.
  • Local-First Privacy: Books, progress, and memory are stored on-device, with remote inference being optional.
  • Broad Format Support: Native support for EPUB, MOBI, AZW3, FB2, PDF, TXT, HTML, CBZ, and CBR without needing format conversion.
  • Focused Reading: Sentence-by-sentence reading mode to reduce distraction.

Related

  • Project
    Swival/swival

    A CLI coding agent optimized for reliability with small and local models, featuring broad provider support and advanced context management.

  • Project
    yeahhe365/Prisma

    A visual deep multi-agent reasoning engine for Gemini and OpenAI-compatible models that decomposes complex queries into parallel expert tasks with a transparent reasoning workflow.

  • Project
    tableau/tableau-mcp

    A suite of developer primitives using the Model Context Protocol (MCP) to integrate Tableau's data and analytics into AI applications.

  • Project
    Meridiona/meridian

    Meridian is a desktop productivity app that automatically reconstructs your workday from screen activity, then uses your connected AI provider to generate daily summaries, standup updates, and Jira worklog drafts ready for you to review and post. It keeps all capture data in an encrypted local database and only sends the data needed for a summary to the AI provider.