MD Analyzer

MD Analyzer is a desktop app that turns Perplexity AI into a Markdown co-pilot. Drag in a stack of .md files, hand the assistant a prompt, and get token-streamed answers grounded in real-time web search — for every file, in batch, with the full conversation kept on disk.

Three modes, one window

• Files — drop one or many .md files and run an AI prompt over each one. Useful for bulk rewriting, summarising, or fact-checking a folder of notes.
• Text — free-form chat with the Perplexity API, no files attached. The same model and conversation history, just without the file context.
• List — define a list of items (e.g. a list of company names) and an optional template like "Write a one-paragraph profile of {{element}}"; the app processes each item separately with a 20-second cooldown between API calls.

What it does

• Perplexity AI integration — pick between sonar, sonar-pro, sonar-reasoning, sonar-reasoning-pro and sonar-deep-research per-prompt.
• Real-time streaming — tokens land in the UI as they arrive, with smooth incremental rendering.
• Prompt queue — chain multiple prompts and let the app run them sequentially with a built-in cooldown so you do not blow through the rate limit.
• Document extraction — when an AI response embeds delimited Markdown blocks, MD Analyzer detects them and offers a one-click save (or auto-save) of each block as its own file.
• Conversation history — optionally include prior messages so each new prompt is context-aware.
• Bilingual UI — Polish (default) and English; the API key, model, temperature and language all persist between runs.

Tech stack

• Shell: Tauri 2.5 + Rust (small binaries, native file dialogs, plugin-store for persistence)
• App: React 19 + TypeScript (strict) + Vite
• Styling: Tailwind CSS with the typography plugin
• State: Zustand
• Rendering: react-markdown + remark-gfm + react-syntax-highlighter
• Localisation: i18next + react-i18next
• AI: Perplexity API (sonar family)