This topic was automatically generated from Slack. You can find the original thread here.
Is there a way to listen to a gmail inbox with pipedream, in order to carry out an action on it? I did find this to be possible with make but I’d like to do the same within pipedream.
Hi Mark, the current supported way that you can do this is by creating your own Gmail app, and creating a Custom OAuth client on Pipedream - you can see our docs here. Once you’ve configured your own custom client, you’ll be able to specify the scopes you need to accomplish listening to your gmail inbox.
That being said - if you’re able to wait a few weeks, we’re in the process of getting our scopes expanded with Google to include a broader set of permissions that would enable exactly what you’re looking for. I estimate that we’ll be able to support this natively in about 3 weeks time.
Hi Micheal, Hi Mark, I’m interested in understanding your limitation, I’ve triggered workflows before on new labeled emails or new attachments received + filter on labeled emails. What is the difference here with the “listening to an inbox” request and “new email received” for example? Why are the custom scopes necessary to answer Mark’s needs? I;m just not following along and think this development might in someway be useful for my workflows too
Hey Diego, if you’re using the sources that you mentioned - you’re using your the Gmail (Developer App), where you’ve configured your own client id and client secret, along with the appropriate scopes necessary to be able to use the triggers and actions for the Gmail (Developer App).
What I’ve mentioned to Mark is that, with the official Pipedream client - our Gmail app, is limited to only Send Mail functionality today, as restricted scopes are necessary for anything beyond this.
Once we get these approved, users will not have to configure their own Gmail client, and can instead use the Pipedream official Gmail client for our components. Does that make sense?
Glad to hear that! If you’ve set up clients in Google Cloud before, it’s not too difficult but I feel like setting up the first app can take a bit of trial and error. Hope the directions were helpful
Hi ! Happy to report that we’ll be releasing this in the next week or so. Just going through a refactor of the components, and I’ll ping you when this is ready - hoping to get this out next week.
I’m so excited!! I couldn’t wait to test this out. My initial reaction is that this will work perfect for an email VA that I coded up. I’m stuck on the last step… I have the label that I’d like to apply, however, I do not know how to programmatically look up the label based on the text value of the suggested label to apply, and also be able to programmatically apply more than one label. Is there a trick to do this? But otherwise this is amazing and will work perfect for my use-case.
FYI I also tried to set up a label name to label id mapper and pass it over as an array of values (tried both quoted and unquoted) thinking that would work, but that didn’t either. I’d really prefer to just pass in the text label name and have the API look up the label id automatically from it.
Ok, I had this run last night for every 30 minutes, with just the single label & hard-coded mapping and it worked! I need to tweak my prompt for more accurate labeling, but it properly tagged all the emails My issues/concerns with labels are still active though. Thanks for this!! Very happy
can you query the google api to ask a list of all existing labels and their id’s so you don’t have to hard-code that list? It will save you time not having to go and edit each time you create a new label? Also, did you try adding multiple “add label” steps one after the other to apply multiple labels?