A small, honest tool for looking at HTTP requests

Why it exists

Integrations fail in the gaps between systems. The documentation says one thing, the provider sends another, and your logs show only what your code managed to parse. A request inspector closes that gap by showing the request itself, before any framework has touched it.

Tapwire is built to do that quickly and without friction: one click for a URL, no account, live updates, and data that cleans itself up.

Principles

• Show the truth. Bodies are stored byte for byte. JSON is re-indented without being re-parsed, so large numbers and odd formatting are shown exactly as sent.
• Keep data briefly. Every endpoint expires and takes its data with it. You can delete anything sooner.
• Private by default. Reading captured requests requires a key that stays in your browser unless you choose to share it.
• Be clear about limits. Tapwire runs on free-tier infrastructure. The limits are published on the documentation page and enforced consistently, rather than hidden behind a promise of "unlimited".
• No ads in the tool. The inspector never shows advertising, and captured request data is never used for advertising or shared with third parties.

How it is built

Tapwire runs on Cloudflare Workers. Each endpoint has its own Durable Object that receives requests, enforces limits and pushes new requests to your browser over a WebSocket. Request history is stored in Cloudflare D1, a SQLite database, and a scheduled job deletes expired data. The interface is written in React and TypeScript, and every content page is prerendered so it loads fast and works without JavaScript for reading.

Get in touch

Questions, bug reports and security reports are welcome. See the contact page.