I’m the Product Lead of Flows and I want your feedback!

My company is using Flows to help our users execute on functionality that is somewhat difficult to do on their own. Our intention is for them to clone our flows and utilize them in their own workspace.

In v 11.29.3, you changed how collections are used within cloned flows. It appears that you now copy the required requests into a new collection.

  1. This causes a disjointed / duplication of collection if our users have already forked our collection.
  2. The collection settings that we have set are lost when this clone happens. So our users have to reconfigure authorizations and variables.

Above all, I wish that users could truly fork our flows (and get updates when we make them)…but I understand how difficult this could be.

At a minimum, I would like for them not to have to continually fix their auth and variables.

1 Like

@gregjancoLogDNA

Thank you for the feedback, I have sent you a message to learn more about the issue you raised!

1 Like

Hi @avionics-specialist3

Looking at your flow, the issue might be using the Start Block in “Any data” mode.

I would recommend toggling off this setting and adding an input to the start block named data with a type of “Record” to see if this addresses the issue you are seeing. (See screenshot below).

This mode (accepting “any” data) is really only intended to be used with a webhook.

1 Like

It would be great to import environment variables, or vault variables, or builtins like {{$guid}} in either an evaluate block or as a general block to be used.

It would also be nice to have pre and post request script blocks, or even full reusable request wrappers to allow for manipulating a request without updating the request in the collection.

Hii,
I am new to community but i do use the postman for quite a while . I just have suggestion While working with query parameters, I noticed the ability to define enum-like values (dropdown options) which is super helpful. I think it would be great if a similar feature could be added for the Body > form-data section. For example, when sending keys like type or category, it would be really useful to predefine possible values as enums, so we can pick from a list instead of typing them manually every time. It’ll reduce human errors and improve consistency during testing.

Thank you.

My suggestion would be that archived snapshots shouldn’t be visible/useable in the dropdown for the flow module block. Once you get more than like 8 snapshots it takes up a lot of the screen and can confuse users. I think if a flow module already has an archived snapshot set on it, that’s fine and you can error out with the “I can’t find this” like normal, but you should then be able to easily see and choose a snapshot that won’t have that error when you check the dropdown at that point.

1 Like

Post & pre flow level scripts are crucial when running tests? Flows really need to mimic the feature set of a collection

I would like to add a script that automatically requests a new access token, if a flow request returns a 401. I can do this at the collection level, no problem, but I cannot see anything in flows that can replicate this behavior?

I would also like a proper report returned, exactly the same as I see, at the end of a collection runner event. I would also like to export the results.

The reason I need to use a flow is that I want my collection folders/requests to be in alphabetical order, and not duplicated. Any requests that are used for housekeeping tasks, I can hide away in a folder called housekeeping. For instance, I might use the API to create two new folders, but obviously I don’t want hundreds of dummy folders to accumulate on our server, so I would like to use the flow to delete these folders with duplicated delete requests [different folder IDs]

But I don’t necessarily want to see these requests front & centre, in my collection. I would like my collection to act more like a library that syncs with our API documentation.

But flows need to be up to the task, which I don’t believe they are yet?

Please could you consider adding these features.

Summary:

  • Post & pre flow level scripts
  • Post & pre flow folder level scripts
  • Post & pre flow action block level scripts
  • Scripts need to be able to access other action request blocks and environment variables, so that we can read and write to the environment
  • Flow runner reports
  • Flow runner report export

HTTP Request blocks in Flows MUST execute Pre-request and Post-response scripts attached to the requests, just like Collection Runner does.

Without this, Flows cannot be used for real-world API testing.

At point 3, I would like to emulate something like this:

// Exclude login folder from auth retry logic

if (pm.info.requestName === "startSession" || pm.info.requestName.includes("login") || pm.info.requestName.includes("no auth required")) {

    // Skip auth retry for login-related requests

    return;

}

// Auto-retry login if session expired (for all other requests)

if (pm.response.code === 401 || pm.response.code === 403) {

    console.log("Session expired for: " + pm.info.requestName);

    

    var retryCount = pm.environment.get("authRetryCount") || 0;

    

    if (retryCount < 1) {

        console.log("Re-authenticating and retrying...");

        pm.environment.set("authRetryCount", retryCount + 1);

        pm.environment.set("retryRequest", pm.info.requestName);

        

        // Run login, then come back to this request

        pm.execution.setNextRequest("startSession");

    } else {

        console.log("Authentication retry failed");

        pm.environment.set("authRetryCount", 0);

        pm.execution.setNextRequest(null);

    }

} else {

    // Success - only reset retry counter if we were in a retry

    if (pm.environment.get("retryRequest") === pm.info.requestName) {

        console.log("Retry succeeded, resetting counter");

        pm.environment.set("authRetryCount", 0);

        pm.environment.set("retryRequest", "");

    }   

}

Thanks

thanks Charles for the feedback. Would support pre/post scripts from collections suffice? Or you want to do these scripts inside the flow itself? I’m not following what you actually want to do with a flow. I will dm you to clarify if that’s ok.