Wrestly

Typically, when doing this sort of work, I’ve always had this habit of running a dummy server that logs out requests and blindly responds to them. This helps me see what is going on, and what external things are either redirecting or posting directly to our APIs. Up to now I have never really done much more than run a server where all requests are console.log’ed and res.send('ok')’ed, but this time around added an SQLite backed so I could see request history, and a nice front end so the payloads were readable. So here is wrestly

Wrestly is a simple http server that will accept any requests and store their body, headers and cookies as well as any query params into a SQLite DB. It allows you to list the requests already made, and look at each one in detail. We don’t really focus on the response, it just gives you a nice 200 and whatever the contents of the original request was. Routes that go via /api/**/* paths will always respond with JSON - which was useful for us - otherwise it just shows Wrestly’s own webpage with the contents of the original request. It’ll parse any JSON or x-www-form-urlencoded bodies included with requests, and it’ll also record the method and url.

You can access the list view at http://localhost:6789/_x and it should show all recorded requests to date. Note that by default requests are stored in memory, so if you quit wrestly and start again it’ll clear saved requests. You can override this by setting a designated sqlite file with the -d eg.

wrestly -d ./requests.sqlite

Combine this with something like ngrok and you have a pretty whizzy environment for working on external webhooked services etc.