Network Explorer

Inspect network requests, parse bodies, and query your traffic with JMESPath.

Network Explorer adds a DevTools panel for inspecting network traffic and querying it. URLs, headers, request payloads, and response bodies are combined into one JSON object per request, which you can filter and transform with JMESPath. FILTER REQUESTS WITH JMESPATH Type an expression and keep only the requests where it is true. It can check any field, including the response body: • status >= `400` • contains(url, '/api/') && request.payload.userId == `7` • response.body.error != `null` QUERY THE WHOLE CAPTURE The collection query runs over every captured request and can return any JSON: a list of failed calls with their error messages, the ten slowest endpoints, or a count of JSON responses. Results update as new requests arrive. READ BODIES AS STRUCTURED DATA • JSON, URL-encoded forms, multipart fields, text, and binary metadata are parsed for you • Collapsible JSON tree with key and value search • Raw toggle to see the original body • Explorer view of the full request object, so you can see which paths to query • Header maps with lowercase names, plus the original raw headers FAMILIAR WORKFLOW • Resource-type filters that match Chrome's Network panel (Fetch/XHR, Doc, CSS, JS, Font, Img, Media, and more) • Record/pause, clear, and preserve log across navigation • Copy or download a single view, a full request, the whole capture, or a query result as JSON • Follows the DevTools light/dark theme, with adjustable font size and font family PRIVATE BY DESIGN Everything runs locally inside DevTools. No server, account, analytics, or host permissions. The extension cannot make network connections, never replays or refetches requests, and never renders response markup. Captured traffic is kept in memory only and is cleared when DevTools closes. HOW TO USE 1. Open DevTools (F12) on any page. 2. Select the "Network Explorer" tab (it may be under the » menu). 3. Reload the page to capture its initial requests. Limits: up to 2,000 requests and 50 MiB of body content per session; bodies over 5 MiB are skipped. Live WebSocket messages and streaming chunks are not captured.