HAR File Browser
A privacy-first HAR inspector: drop in a .har file and explore the network waterfall, headers, payloads, cookies, and timings — parsed entirely in your browser, never uploaded.
Proposed stack
- Next.js
- TypeScript
- Tailwind CSS
- Client-side parsing
Goal
Give developers a fast, private way to inspect HAR (HTTP Archive) captures without uploading sensitive traffic to a third-party service.
Overview
HAR File Browser loads an HTTP Archive (.har) file — uploaded or pasted — and renders the captured network traffic as an interactive waterfall: each request with its method, URL, status, type, size, and timing bars.
Click any request for the full detail: general info, request & response headers, query string, payload, response body, cookies, and the timing breakdown (DNS, connect, SSL, send, wait, receive).
Everything runs in the browser — HAR files contain cookies, auth tokens, and request bodies, so nothing is ever uploaded, stored, or logged. Filters and summary stats round it out.
Architecture
HAR File → Client-side Parser → Waterfall & Filters → Request Detail
Key features
- Upload or paste a .har file (drag & drop).
- Interactive request waterfall with timing bars.
- Full request detail: headers, query, payload, response, cookies, timings.
- Filters: search, method, status class, resource type.
- Summary stats by count, size, time, type, and status.
- 100% client-side — nothing is uploaded, stored, or logged.
Technical challenges
- Parsing varied HAR exports (Chrome, Firefox, Charles) consistently.
- Rendering a readable waterfall and deep request detail without clutter.
- Keeping everything client-side so sensitive captures never leave the browser.
How it's built
- 1Next.js + TypeScript + Tailwind, with the HAR parsed in-memory in the browser.
- 2A waterfall timeline plus a detail panel driven by the parsed entries.
- 3On-the-fly filtering and summary stats; deployed to Vercel.
Interested in this work?
Happy to talk through the design, trade-offs, and implementation plan.