Memoire

Product Introduction

1. Definition: Memoire is a local-first, distraction-free desktop writing application built for long-form content creation. Technically, it is a cross-platform (macOS, Windows, Linux) Markdown editor with a Tauri architecture (Rust core, SvelteKit front end) that prioritizes data privacy and offline functionality.
2. Core Value Proposition: It exists to provide writers with a fast, private, and keyboard-centric environment that eliminates digital distractions and cloud dependency. Its primary value is enabling a deep "flow state" for drafting while ensuring user data remains under their complete control with plain-text file storage.

Main Features

1. True Distraction-Free Canvas & Focus Mode: The editor implements a meticulously designed typographic system with optical margin alignment, a 68-character line measure, and a curated type scale for extended reading. The "Focus Mode" (activated by Cmd + .) is a comprehensive state that engages full-screen, collapses all UI panels, and employs dynamic dimming, highlighting only the active paragraph. This is powered by custom CSS transitions and a typewriter scrolling algorithm that keeps the caret centered without disrupting mouse selection.
2. Local-First Architecture with Optional Sync: Memoire's foundational technology is its local-first data model. All documents are stored as standard Markdown (.md) files in a user-designated folder on the local disk, requiring no internet connection. Optional sync is implemented via two distinct backends: a Git client for committing to GitHub repositories (enabling full version history diffs) and a direct sync engine for Dropbox, which performs content-hash comparisons to only transfer changed files. Authentication tokens are stored in the OS keychain.
3. Intentional, Offline-Capable AI Assistant: Unlike always-on AI writing tools, Memoire's assistant is disabled by default. When enabled, users choose between running quantized GGUF models (like Qwen 2.5 or Llama 3.2) locally on their hardware (no data transmission) or connecting to OpenRouter with a personal API key. The AI functions contextually via a selected text menu for discrete actions like "Polish," "Tighten," or "Change tone," presenting edits as a clear diff for approval.
4. Keyboard-Driven Command Palette & Workflow: The entire application is navigable via keyboard shortcuts. The unified command bar (triggered by :) uses fuzzy matching on both command names and descriptions. Core writing workflow shortcuts include Cmd+N for instant new document creation (~40ms), Cmd+U for a non-destructive, branching version history graph, and Cmd+Shift+E for export workflows to formats like EPUB, DOCX (with manuscript formatting), and PDF.

Problems Solved

1. Pain Point: Cloud-based writing apps create vendor lock-in, potential privacy issues through telemetry, and disrupt workflow with constant connectivity requirements and distracting, omnipresent AI suggestions.
2. Target Audience: The primary user personas are serious writers who value ownership and privacy, including novelists, academic researchers, journalists, technical writers, and bloggers. Secondary users are developers and technologists who appreciate plain-text workflows, Git integration, and open-source software.
3. Use Cases: Essential for drafting novels or long-form essays offline (e.g., on a plane); maintaining a private, future-proof writing archive in Markdown; integrating writing directly into a developer's existing Git-based version control and backup system; and achieving deep focus without UI clutter or unsolicited AI interruptions.

Unique Advantages

1. Differentiation: Compared to cloud-first competitors like Notion or Google Docs, Memoire offers complete data ownership and offline reliability. Versus other Markdown editors like Obsidian or Typora, it distinguishes itself with a stricter, more opinionated focus on the drafting experience (e.g., its unique Focus Mode implementation and session-based word goals over streaks) and its built-in, optional local AI integration.
2. Key Innovation: Its "branching version history" is a significant innovation. Instead of a linear undo stack, each save creates a persistent node. Users can backtrack (Cmd+Opt+Z) and create alternate drafts, preserving all paths in a visual graph stored as a local file, enabling non-destructive exploration of different narrative or editorial directions.

Frequently Asked Questions (FAQ)

1. Is Memoire really free and open source? Yes, Memoire is completely free to use and its source code is publicly available under the permissive MIT license on GitHub, allowing for transparency, community auditing, and contributions.
2. How does Memoire's local AI work without an internet connection? It uses quantized GGUF model files (typically 1.5GB - 8GB in size) that are downloaded once and run entirely on your computer's CPU/GPU via local inference engines like llama.cpp, ensuring no data ever leaves your device.
3. Can I use Memoire with my existing folder of Markdown files? Absolutely. You simply point Memoire's "content directory" setting to your existing folder. It will index the .md files immediately, making them available in your binder without any import or conversion process.
4. What happens if there's a sync conflict between my laptop and desktop? With Git sync, you resolve conflicts using standard Git merge strategies. With Dropbox sync, Memoire employs a safe "copy-rename" strategy: it downloads the conflicting remote file but saves it with a " (Dropbox conflict)" suffix, preserving both versions locally for manual resolution.
5. Does the Focus Mode work on multiple monitors? Yes, the Focus Mode (Cmd + .) engages full-screen on the monitor where the Memoire window is located, providing a truly immersive, distraction-free writing canvas on that display.