ParseKit

Documents → clean Markdown, on your Mac.
Menu bar · offline · no accounts · no API keys

↓ Download for Mac (Apple Silicon)  ·  Install guide  ·  Releases

See it in action

Convert files locally	See completed batches
Make ParseKit yours	Check file support

---

What it does

Drop PDFs, Word, Excel, PowerPoint, or images into the menu-bar app. ParseKit converts them locally to Markdown, plain text, or JSON — ready for ChatGPT, Claude, notes apps, or RAG pipelines.

Nothing is uploaded. Your files stay on your Mac.

You get	Details
Local conversion	PDF, Office, images → .md / .txt / .json
OCR	Scanned pages become real text
Finder Quick Action	Right-click → Parse to Markdown
Hotkey + clipboard	Global shortcut (⌃⇧M)
CLI	parsekit convert for scripts and agents
Updates	In-app install when a new version ships

PDF works out of the box. Word/PowerPoint need LibreOffice; images need ImageMagick (brew install imagemagick). Settings → File Support shows what’s missing.

Install (1 minute)

1. Download the DMG
2. Open it → drag ParseKit to Applications
3. Open from Applications → look for the icon in the menu bar (top right)

First launch may be blocked by Gatekeeper (the app isn’t notarized yet). Fix once:

xattr -cr /Applications/ParseKit.app

Or right-click → Open, or use Settings → General → Copy fix command. Full steps: docs/INSTALL.md.

How people use it

• Paste into AI — convert a scan or contract, copy clean Markdown into Claude/ChatGPT
• Batch prep — drop a folder of client docs, convert once, feed the output folder into a project or RAG index
• Agent workflow — parsekit convert spec.pdf --out /tmp/spec.md before loading a PDF into context

More: skills/parsekit/SKILL.md · AGENTS.md

Privacy

Files, OCR, and conversion never leave your Mac. No telemetry. Network only for optional app updates and first-time OCR language packs.

Product metrics (local-first)

Metric	Definition	Storage
Activation	First batch with at least one successful file	activatedAt in the Tauri settings store
Tokens saved	Estimate: raw extract vs cleaned Markdown (~4 chars/token)	On-disk token stats
Today / all-time	Calendar-day and lifetime sums of token events	Same
Success rate	Successful vs failed files across the last 20 jobs	jobOutcomesV1 in the settings store (counts only)

Stored only on this device. Never uploaded. See docs/TASKS.md.

Develop

git clone https://github.com/harshabala/parsekit.git
cd parsekit
npm install
npm run build:sidecar
npm run tauri dev

Checks: npm test (Vitest), npm run check (svelte-check), npm run build (frontend bundle). Requires Node, Rust, and Tauri prerequisites.

Release process: docs/RELEASING.md · Benchmarks: docs/benchmark-results.md

Credits

Created by Harsha Balakrishnan.
Powered by LiteParse · Tauri · Svelte
License: Apache-2.0