DotMD

About DotMD

DotMD is a native Mac Markdown application where users review documents written by AI agents. It presents the review process as a pull request interface, allowing line-by-line commenting, agent-led fixes over the Model Context Protocol (MCP), and per-change accept/reject decisions. The app also functions as a standalone Markdown reader and editor with support for Mermaid diagrams, math notation, code blocks, tables, front matter, version history, search, and PDF export.

Review

DotMD targets a specific friction point in AI-assisted development: the back-and-forth of reviewing agent-generated specification files. Instead of quoting paragraphs in a chat window and losing track of revisions, the app anchors comments to exact lines and displays proposed changes as red/green diffs directly in the document. The tool runs entirely on-device with no cloud account required, and its MCP server listens only on localhost.

Key Features

• Pull request-style review for Markdown. Select a passage, leave a comment, and an AI agent connected over MCP receives the quoted text, line range, and comment. The agent edits the file on disk, then DotMD renders the fix as a diff next to the original comment for acceptance or rejection.
• Local MCP integration. The built-in MCP server runs on localhost. Agents get only the specific passage and line range cited in a comment, not the full document, though they can read more context from the file as needed.
• Version history per file. Every document keeps its own change record, making it possible to review how a spec evolved across multiple agent revisions.
• Multi-format reading and editing. The editor handles Mermaid diagrams, LaTeX math, syntax-highlighted code blocks, tables, and front matter. Folder-wide search and PDF export are included.
• Fully local operation. No account, no cloud sync. All files and the MCP server stay on the user's Mac.

Pricing and Value

Reading Markdown files in DotMD is free with no time limit. Pro features-editing, the review and commenting workflow, MCP integration, and version history-cost $19.99 per year (approximately $1.67 per month). A 3-day free trial is available for the Pro tier. The maker has also mentioned a $50 lifetime license, though feedback from early users suggests this price point is being evaluated. DotMD sells directly through the App Store and via Paddle for purchases outside the App Store, with RevenueCat handling subscription and purchase sync across Macs.

Pros

• Anchors AI agent revisions to specific lines, so users see exactly what changed per comment rather than scanning a rewritten file.
• Runs entirely on-device with a localhost-only MCP server; no account creation or internet connection required for core functionality.
• Reading mode is free indefinitely, which lets users evaluate the editor and viewer before committing to Pro.
• Supports technical document formats that plain Markdown editors often skip, including Mermaid charts and math notation.
• Per-file version history creates an audit trail of agent edits, which is useful for iterating on specs across multiple sessions.

Cons

• The review workflow depends on AI agents that support MCP. Users whose agents or coding tools lack MCP integration won't be able to use the core pull request feature.
• DotMD is a Mac-only native app. Teams or individuals on Windows or Linux cannot run it, and there is no web version.
• The tool is not well suited for users who don't generate documents through AI agents. If you write and edit your own Markdown files manually, the pull request review loop adds steps without a clear benefit over a standard editor.

DotMD fits a workflow where AI agents produce frequent spec drafts and the human's job is to review and approve incremental changes. It makes the most sense for developers who already use MCP-compatible agents like Claude Code or Codex and want a structured review layer between agent output and final document. For teams outside the Apple ecosystem or those whose agents don't speak MCP, the app's constraints will likely outweigh its diff-based review approach.