Show HN: Marky – A lightweight Markdown viewer for agentic coding
TL;DR Highlight
This macOS desktop app allows you to open Markdown files generated in real-time by AI agents like Claude directly in the terminal and view them with live rendering. It simplifies the document review process in AI-powered development workflows.
Who Should Read
macOS developers who want to quickly check planning documents, READMEs, notes, and other files generated while coding with AI agents (Claude, Cursor, etc.). Especially those who want to conveniently manage a large number of Markdown files output by an agent.
Core Mechanics
- Marky is a macOS-exclusive desktop app that instantly renders Markdown files or folders with a single command in the terminal: `marky FILENAME` or `marky FOLDER`.
- It supports live reload, so the viewer screen automatically updates when Claude or an editor modifies the file without requiring a separate refresh. You can see the content in real-time as the AI writes the document.
- If you register a folder as a workspace, a file list appears in the sidebar like Obsidian, and it is automatically restored even when the app is restarted.
- You can fuzzy search all files in registered folders with Cmd+K command palette, and it internally uses the nucleo library.
- Code blocks are rendered with Shiki (the same syntax highlighting engine as VS Code), equations with KaTeX, and diagrams with Mermaid as SVG rendering. It supports the entire GFM (GitHub Flavored Markdown).
- HTML rendering is sanitized with DOMPurify, so you can safely open untrusted Markdown files.
- The tech stack is Tauri v2 (Rust-based desktop app framework) + React + markdown-it, and it can be installed with Homebrew or built directly from source.
Evidence
- There was feedback that the term 'native' could be misleading. While Tauri uses the system WebView, it's an exaggeration to call it 'JavaScript native' since it's ultimately based on WebView.
- Several people mentioned having created similar tools themselves. vantageapp.dev, sdocs.dev, seams, hyprmark, mdreader, and other similar projects were shared in the comments, demonstrating that the boom in AI agent development has led to a simultaneous increase in people trying to solve this problem.
- The question 'What's the difference between this and Obsidian?' was raised. Obsidian also supports folder views and Markdown rendering, and the distinction was unclear. There were no clear rebuttals, and the comment did not lead to further discussion.
- There was also the opinion 'VS Code + Markdown preview is almost the same functionality.' However, this comment also included a specific need: integration with git and highlighting changes made by the agent.
- A UX inconvenience was reported: the text size cannot be adjusted. Along with feedback that the sidebar column width cannot be adjusted, there were real-user opinions that the default size is small and strains the eyes.
How to Apply
- Have an AI agent like Claude or Cursor write a planning document or spec file in Markdown, then open the corresponding folder in a separate terminal with the command `marky ./plans/` to view the content live as the agent updates the files.
- In projects where an agent generates a large number of Markdown files, open the project root with `marky .` and quickly find the desired file with Cmd+K fuzzy search. It is possible to quickly navigate even if more than 75 files are generated.
- Since it supports Mermaid diagram and KaTeX equation rendering, you can prompt the AI to write system architecture or flows in Mermaid code blocks and instantly visualize them in Marky.
- Because DOMPurify sanitization is applied, you can safely render and view Markdown files received from untrusted sources without the risk of XSS when checking documents received from outside.
Code Example
# Installation (Homebrew)
brew install marky
# Open a single file
marky README.md
# Open the entire folder as a workspace
marky ./docs/
# Open the current directory
marky .Terminology
Related Papers
ChatGPT for Google Sheets Exfiltrates Workbooks
Google Sheets용 ChatGPT 확장 프로그램이 간접 프롬프트 인젝션 공격에 취약해, 단 하나의 시트에 숨겨진 악성 명령만으로 계정 내 워크북 전체가 외부로 유출될 수 있다는 보안 연구 결과가 공개됐다.
LinTree: Improving LLM Reasoning with Explicitly Structured Search Histories
LLM의 추론 트레이스에 부모 포인터(parent pointer)만 추가해도 탐색 성능과 효율이 크게 올라간다.
Diagnosing Failure Modes of Shared-State Collaboration in Resource-Constrained Visual Agents
4B~8B 소형 비전 모델에서 공유 메모리(화이트보드) 기반 멀티에이전트 협업이 오히려 성능을 떨어뜨리는 이유를 분석한 연구.
Show HN: Open Envelope – an open schema for defining AI agent teams
여러 AI 에이전트가 팀처럼 협업하는 구조를 벤더 중립적인 오픈 스키마로 선언적으로 정의할 수 있게 해주는 프로젝트로, 멀티 에이전트 오케스트레이션의 표준화를 시도한다.
Unifying Temporal and Structural Credit Assignment in LLM-Based Multi-Agent Prompt Optimization
여러 AI 에이전트가 협력할 때 '어느 라운드의 어느 에이전트'가 실패했는지 정확히 짚어내서 그 프롬프트만 고치는 최적화 프레임워크
Show HN: Ktx – Open-source executable context layer for data agents
AI 에이전트가 회사 데이터 웨어하우스를 정확하게 쿼리할 수 있도록 시맨틱 레이어, 메모리, 비즈니스 지식을 자동으로 구축해주는 오픈소스 도구다. 기존 에이전트가 매번 웨어하우스를 재탐색하거나 잘못된 메트릭 로직을 임의로 만들어내는 문제를 해결한다.