57 comments

  • ajnin 16 hours ago

    Hum I'm definitely not the target for this but I don't see the value to obfuscate all the info in an UUID, I'd have kept things simple and stored things into an URL fragment, that way it's possible to operate client-side only and nothing gets leaked to a server I don't know much about like headers or whatever tokens or API keys could be passed in an URL.

    • themanmaran 8 hours ago

      I see a fair bit of utility here. We have an API product, and (despite having pretty clear docs) I almost always have to send people curl requests they can copy and run.

      So I see this as similar to having a sandbox built into your docs page. Except I can customize a request and send it directly to a user. The only missing piece is the authentication part. Since I wouldn't want to embed an api key in this link.

      • darubramha 5 hours ago

        Totally with you on that—your use case is exactly what I had in mind when I built uncurl.dev.

        I kept finding myself sending curl requests in Slack or email, and it felt clunky—especially when non-devs or support teams needed to test something quickly. uncurl.dev started as a way to make that share-and-execute process more visual and frictionless.

        For the auth part—embedding API keys in payload is a no-go in most cases. For now, sending out auth headers separately for them to fill in themselves is what I did within my workplace.

        I'm exploring a couple of ideas to help with this:

        Team-scoped secrets: For logged-in users or teams, saving common auth headers that aren’t part of the shared link.

        One-time, encrypted secrets: The link works once and destroys the sensitive payload after execution.

    • aae42 16 hours ago

      could also just do the request in javascript instead of needing a (presumably hosted) sandbox

      • TheDong 15 hours ago

        curl supports many things javascript doesn't. Like http proxies, tls versions, and half the other flags

      • ultrafez 15 hours ago

        There are many types of request that cannot be made with client-side JS alone, but for those that can, the ability to send those requests client-side would be handy.

        • markerz 12 hours ago

          I think that 99.9% of CURL commands are copied from Chrome/Firefox's network inspector and are the simple "client-side JS" types.

          I also think it's weird to be so willing to let people run arbitrary CURL commands from your platform, without any billing or account verification. It feels ripe for abuse.

          • jasir 12 hours ago

            Even for commands that use the subset of cURL features that fetch supports, most requests to other domains (cross origin requests) wouldn't work anyway because the responses won't have the CORS[0] headers to allow being accessed from arbitrary websites. So running it client side would be infeasible for most requests.

            [0]: https://developer.mozilla.org/en-US/docs/Web/HTTP/Guides/COR...

            • darubramha 5 hours ago

              Exactly — this was one of the core constraints I ran into early on.

              CORS was a blocker for client side requests, I have a separate branch where this is integrated, maybe will add it alongside server side execution to let the person creating the curl decide whether they can execute on browser or server side.

  • markerz 12 hours ago

    Hey OP, your DELETE curl endpoint is unauthenticated! I can't DM you on HN and there's no contact on your website, so sorry for the public security disclosure. :(

    • mzronek 12 hours ago

      That's part of the vibe in vibe coding.

    • darubramha 5 hours ago

      Hey, thank you so much for catching this and calling it out , will take this up and fix it!

      Really appreciate you taking the time to look and let me know (even if it had to be in public). I have added a github repo for filing bugs (https://github.com/uncurl/uncurl-support) in the docs page :)

  • treesknees 15 hours ago

    This is why in our codebase we have a rule to not use short options/flags for called commands like curl. And if there is only a short option available, it must be explained in a code comment.

    • xrisk 13 hours ago

      Why do you call curl via the binary instead of using libcurl?

      • treesknees 10 hours ago

        As others stated, shell scripts and some other automation that have to call the cli tool directly, or where we need to build out the cmd and execute it on a remote node.

        We use libcurl and pycurl where it makes sense. This rule for cli options extends to other binaries as well, some that don’t offer libraries like curl does (think closed source firmware tools or ancient homegrown cli tools.)

      • markerz 13 hours ago

        Docker files too, basically just bash scripts with a different syntax

      • notpushkin 13 hours ago

        Bash scripts?

  • trollied 12 hours ago

    Flagged this because it is a security clusterfuck.

    • darubramha 5 hours ago

      Fair. I appreciate the honesty — even if it's a bit brutal :)

      Security is a top priority for this project, and I'm actively working to tighten things up. This initial version was launched to validate the concept, and admittedly, there were oversights (including an unauthenticated DELETE endpoint that was highlighted).

      If you're open to it, I'd love to learn more about what you'd want to see from a security standpoint in a tool like this. I'm building in public and happy to be corrected where needed.

      Thanks again for keeping things real.

  • Tabular-Iceberg 16 hours ago

    It seems very particular about what curl options it supports. I keep getting “Please enter a valid curl command” no matter what I do.

    Maybe the only solution is to somehow extract the actual command line parser from curl itself.

  • benoitg 14 hours ago

    The Jetbrains suite of IDEs have this handy feature : if you copy a curl command into an HTTP scratch file, it is automatically converted to the HTTP equivalent, which is IMHO much more readable.

    Your project looks very cool though, and expands on the share aspect of the Jetbrains feature, very interesting!

    • darubramha 5 hours ago

      Appreciate the kind words—and the comparison!

      uncurl.dev kind of grew out of that same spirit, but with the goal of making the output shareable and executable in a browser, especially for folks who might not have an IDE set up or are outside the usual dev loop (PMs, etc.)

    • hk1337 10 hours ago

      Paw (Now RapidAPI) has a nice feature that generates a http request. The curl plugin and http plugin are the main ones I use but there's several others to generate the code in other languages too.

    • _kidlike 12 hours ago

      my thoughts exactly!

      Lately I feel like a lot of people think they are finding gaps around developer experience, but it's only because they don't know the right tools that already exist...

  • h1fra 16 hours ago

    Interesting. Default behavior could be improved. I blindly pasted a curl, except showing my curl it didn't make any headers modifiable. It also didn't redacted the Authorization header. Also there is no way to delete a page.

    • markerz 12 hours ago

      FYI, you can delete anyone's CURL (including your own if you were unauthenticated) with the following curl:

      https://uncurl.dev/curl/78ab4bf5-34e8-45a0-b3b1-32dd6aa7e360

      or this command

          curl -X DELETE "https://uncurl.dev/api/curls?id=051606b5-49c8-4f14-9689-4d424f71d331"
      
      Looks like deletes are unauthenticated.
      • darubramha 5 hours ago

        Haha love that you shared the curl with the uncurl.dev url!

        Yes, delete is unauthenticated as highlighted, will be working on a fix for this. And you can delete any API if it is created as a logged in user.

    • serial_dev 15 hours ago

      Whether redacting the auth header is the best choice can be determined on a case by case basis, so I don't think it should redact by default. A big scary warning would definitely make sense, though!

    • byearthithatius 13 hours ago

      Exact same thing happened to me. Had to reset my HN user cookie because accidentally pasted my downvote curl command.

  • lenkite 13 hours ago

    Feels like a security nightmare - this is far better distributed as a local desktop UI rather than one hosted.

    • byearthithatius 13 hours ago

      It 100% is because he just keeps your curl and you can't even delete it. He just keeps it forever for himself even if you made a whoops. Already accidentally uploaded my HN user cookie and had to reset it. JFC.... -_-

      • darubramha 5 hours ago

        I have implemented auto-purge all curls that are created without login will get auto deleted after 30 days of non-usage.

        You can absolutely delete your curl if you have created as a logged in user.

      • thomasfromcdnjs 5 hours ago

        lmao I did similar.

  • niek_pas 16 hours ago

    Looks cool! One bug I found:

    `curl www.google.com` works using 8.7.1 on macOS, but I get "Please enter a valid curl command" on your website.

    • darubramha 5 hours ago

      :) the curl command parser expects some flags of GET, POST, stupidly overlooked, will get it fixed!

  • flipperto 14 hours ago

    While it looks good and even possibly useful, it seems to be a great way to leak sensitive cookies (especially since "copy as cURL" is so easy on the browser's network tab).

    I would 100% forbid its use in a company environment and I would encourage people in general not to use it for any non-trivial use case.

    • byearthithatius 13 hours ago

      Thank you. I literally already accidentally uploaded a cookie that I now need to reset because the website doesn't let me delete the curl -_-

  • fitsumbelay 13 hours ago

    This is a pretty cool project.

    One thing: it's rejecting dict lookups as invalid URL, eg. `curl dict://dict.org/d:failure:fd-eng-fra`

    I'm checking first here whether I missed something in the docs about not supporting DICT before I add issue to the GH repo

    • darubramha 4 hours ago

      You didn’t miss anything in the docs. Right now uncurl.dev only supports http/https (and technically ftp, though it’s untested). Protocols like dict://, smtp://, etc. aren’t parsed or handled correctly yet, which is why you’re seeing that “invalid URL” error.

      I hadn’t actually considered dict:// usage, I see the bug report as well, thanks, will see if I can include it.

  • byearthithatius 13 hours ago

    This is just waiting for people to leak cookies oh my lord....

  • ustad 14 hours ago

    Hey, that looks great.

    Could you describe more about the docker sandbox that you have? I am especially interested in the network restrictions.

    • darubramha 4 hours ago

      The sandbox is a lightweight Alpine-based container, it runs as a non-root user for security, it has minimal dependencies installed (curl, bash, coreutils)

      The container has restricted outbound access—only HTTP/S requests are allowed. It runs inside an isolated network namespace with no access to the host network or other infrastructure components. There's no inbound access, and the container can't receive unsolicited requests from the outside world.

      The sandbox container can only communicate with other containers in the same network, the main application container and sandbox container are on the same network, allowing them to communicate.

  • VWWHFSfQ 16 hours ago

    This would be useful if it was client-side only. I only very rarely have curl commands to run that don't also include some stuff like cookies and tokens, which I'm not sending to someone's server so they can run curl for me.

    • byearthithatius 13 hours ago

      Multiple people already did and are confirming they can't even delete it. This website is literally a security issue itself.

      • darubramha 4 hours ago

        Hey, you can absolutely delete any curls created if done from a logged in user, for others it will get auto-purged in 30 days of non-usage.

    • cgannett 15 hours ago

      I cant even get the thing to work

  • polishdude20 15 hours ago

    Why the need for an account to execute? Are you executing the command on behalf of the user on your server? Is it possible to just do it locally in browser?

    • darubramha 4 hours ago

      Yes — executions are done server-side, inside a resource-limited, sandboxed Docker container. That’s why login is required: to prevent abuse, rate-limit usage.

      I have a feature working to allow users browser side execution, but as others have also pointed out CORS is a big blocker for client side execution not working for all APIs.

    • zamadatix 15 hours ago

      The post mentions the server is used during the optional execution. My guess as to why would be primarily because there are plenty of curl commands which are impossible to execute via the browser due to browser restrictions and secondarily because the execution is literally running curl, not trying to approximate the equivalent action.

      Local execution could still be a handy feature for at least the most common basic commands though, but it'd start to have to wade into "explaining a lot to the user why this isn't really what the result might look like" territory.

      • VWWHFSfQ 14 hours ago

        It actually looks like the server is used regardless. It's sending the full curl command to the backend even if not executing it.

        • zamadatix 10 hours ago

          Well yeah, it's generating short links. I mean to highlight the server runs curl in a sandbox against the 3rd party host on behalf of the user, not just that the server and client talk to each other as you use the page.

  • byearthithatius 13 hours ago

    I already accidentally uploaded a cookie that I now need to reset because the website doesn't let me delete the curl -_-

    • dwrowe 12 hours ago

      There is both this post about not being able to delete a curl, and other posts about being able to delete curls unauthenticated. Looks like you should be able to issue a DELETE curl request, to delete your curl.

    • markerz 13 hours ago

      Definitely reset your password/account even if you can delete it from their website. Otherwise, you are trusting a stranger with no privacy policy or reputation.