Review Markdown.
Feed the LLM.
Stop shipping raw strings. Structural review for AI development plans, PRDs, and technical specs. Optimize your Markdown for LLM consumption and human clarity.
Core utilities
Claude Code skill
Deploy our reviewer as a MCP skill directly into Claude. Let AI audit its own documentation workflow.
- $ git clone https://github.com/juanallo/md-review.git
- $ cd md-review
- $ ./install-skill.sh ~/.agents/skills
Use ~/.claude/skills or ~/.cursor/skills if that is where your agent loads skills from.
VS Code extension
Integrated line-by-line commenting and collaborative feedback system directly in your IDE.
$ coming soon...
The workflow
01
Upload MD
Drag any markdown file or paste raw text into the reviewer engine for parsing.
02
Comment
Anchor feedback to specific line numbers. Tag collaborators or LLM prompts for clarification.
03
Export
Generate a 'Review Manifest' that can be ingested by Claude, ChatGPT, Gemini, or any other LLM.