Jot 2.0

Product Introduction

Definition: Jot 2.0 is a local-first markdown notes inbox utility explicitly engineered for macOS. Categorized under developer productivity tools, lightweight scratchpads, and rapid-capture utilities, it acts as a low-friction conduit between selected desktop content, speech dictation, screen annotations, and a locally owned file system. Rewritten entirely from Electron into native Rust, the software footprint is dramatically reduced from 675 MB down to 43 MB on Apple Silicon architectures, establishing an ultra-efficient resource footprint for performance-conscious Mac users.

Core Value Proposition: Jot 2.0 exists to eliminate friction in the information-capture workflow, enabling users to grab text, images, clipboard screenshots, or audio notes via a single system-wide keystroke and store them instantly as plain-text markdown files inside a user-owned directory (~/Documents/Jot). By prioritizing a local-first philosophy—featuring zero mandatory user accounts, zero cloud synchronization requirements, and local AI model integration—Jot 2.0 ensures complete data sovereignty while integrating tightly with modern AI coding assistants via built-in CLI and Model Context Protocol (MCP) server support.

Main Features

Instant Keystroke Capture (Shift Shift): Jot 2.0 intercepts selected text, embedded images, or active clipboard screenshots using a double-tap of the Shift key. This action instantly parses the content, assigns unique metadata, and writes a structured markdown file (YYYY-MM-DD-id.md) into the designated local directory without opening a heavy application window.

Local Voice Dictation & AI Transcription: By holding the fn key, users can dictate text directly into any active macOS application. Jot 2.0 supports dual-stage voice processing where transcription and text cleanup are handled independently. Local speech models include Parakeet, Apple Speech (macOS 12+), Apple Intelligence (Apple Silicon, macOS 26+), Ollama, and LM Studio. For cloud-centric workflows, users can supply API keys (stored securely in the macOS Keychain) for OpenAI, Anthropic, Gemini, Groq, Mistral, OpenRouter, or the Vercel AI Gateway.

Screen Annotation & Quick Notes (⌃⌥A & ⌃⌥N): Users can capture and mark up specific screen regions using ⌃⌥A (or draw a new canvas via ⌃⌥D), copy the result to the clipboard, or save it directly to the local inbox. Meanwhile, ⌃⌥N initiates a headless quick-note entry. The interface is optimized around a keyboard-first command palette (⌘K), allowing for rapid note merging, searching, archiving, and deletion.

AI Agent Integration (CLI & MCP Server): Jot 2.0 bridges raw notes and development workflows through a native command-line interface and an integrated Model Context Protocol (MCP) server (jot mcp). This architecture allows AI coding agents like Claude Code, Codex, Cursor, T3 Code, Windsurf, and Zed to list, search, read, add, finish, archive, and tag local jots directly through chat interfaces without requiring manual copy-pasting.

Problems Solved

Pain Point: Knowledge workers, developers, and researchers frequently suffer from "context fragmentation," where valuable snippets, error logs, thoughts, or meeting excerpts are trapped across web browsers, notification centers, and bloated note-taking apps that require active window switching, internet connectivity, or paid cloud subscriptions. Jot 2.0 solves this by abstracting note-taking down to a subconscious, keyboard-driven reflex that operates completely offline.

Target Audience:

• Software Engineers & Developers: People working within terminal environments and AI coding assistants (Claude Code, Cursor, Zed) who need to capture error traces, CLI outputs, or documentation snippets and instantly expose them to local LLMs via MCP.
• Writers & Researchers: Academics and content creators collecting multi-source citations, web selections, and clipboard assets into persistent, non-proprietary plain-text files.
• Power Mac Users: Minimalists and privacy advocates who reject electron-based memory hogs, subscription fatigue, and forced cloud storage syncing.

Use Cases:

1. Capturing a code stack trace from an IDE, hitting Shift Shift, and immediately sending it to Claude Code via right-click or ⌘⇧S to debug.
2. Dictating meeting action items post-conference using local Parakeet speech recognition while keeping sensitive business data completely off external servers.
3. Snip-annotating a UI bug using ⌃⌥A and filing it directly as a markdown task card within a local repository-linked folder.

Unique Advantages

Strengths & Limitations (Pros & Cons):

• Pros: Extremely lightweight native Rust build (43 MB); total data ownership via plain markdown files; zero accounts or forced subscriptions (one-time license); robust developer extensibility via CLI and MCP server; powerful local and cloud AI transcription flexibility.
• Cons: Strictly limited to macOS (macOS 12 or newer required; specific features require newer OS versions or Apple Silicon); lacks native cross-device cloud synchronization out-of-the-box (relies on user-managed sync solutions like iCloud Drive or Git if multi-mac synchronization is desired); learning curve for mastering keyboard shortcuts and agent piping workflows.

Key Alternatives & Differentiation:

• Obsidian / Logseq: While both are powerful local-first markdown knowledge bases, they are full-fledged workspace applications focused on graph linking and heavy vault management. Jot 2.0 positions itself strictly as a zero-friction capture inbox that sits upstream of tools like Obsidian or agentic coding environments.
• Apple Notes / Quick Note: Apple's native solutions lock notes inside proprietary databases, lack deep programmatic command-line and MCP server access for AI agents, and lack flexible bring-your-own-key local AI dictation pipelines.
• Raycast / Alfred Snippets: While excellent application launchers, they lack dedicated persistent markdown inbox archiving, screen annotation markup routing, and native agent context integration out of the box.

Frequently Asked Questions (FAQ)

Is Jot 2.0 a subscription service? No. Jot 2.0 operates on a one-time, perpetual license model starting at $9 for a single Mac during the launch phase, scaling to 3-Mac and 10-Mac bundles. There are no recurring subscription fees, and software updates are included.

Where are my notes stored, and can other applications read them? Every jot is saved as an individual plain-text markdown file inside ~/Documents/Jot/. Because they use standard markdown and frontmatter, any text editor, note-taking app, or AI coding assistant (via CLI or MCP) can read, search, and parse them directly without proprietary lock-in.

Does Jot 2.0 require an internet connection or a user account? No user account is required to use the application. Jot 2.0 is fundamentally local-first; all core capture, file generation, and local speech transcription (using engines like Parakeet, Ollama, or LM Studio) operate entirely offline. Cloud transcription or cleanup APIs are optional and require user-provided API keys stored securely in the macOS Keychain.

How does the MCP server integrate with AI coding tools? Jot 2.0 includes a built-in Model Context Protocol (MCP) server executable. By configuring your client (such as Claude Code, Cursor, or Codex) using the provided setup command, your AI assistant gains secure, direct read, search, and tagging access to your local jot inbox to contextualize coding tasks.