Add custom pending transactions via API
D
Daniel Prilik
I'm working on a little project to sync splitwise transactions to lunch money (https://github.com/daniel5151/splitwise-lunchmoney), and it'd be cool to have a credit-card style "pending/posted period" for API-imported transactions, similar to what plaid-imported transactions have.
i.e: transactions could be set to appear "pending" for a few days (e.g: while friends make quick edits), and then transition into a "posted" state.
My tool is totally workable without this feature... but form a UX perspective - it'd be a great way to visually indicate that the transaction is still "in flux"
Log In
S
Sylvain Francois
Also interested in this feature. I’m building an integration for Fortuneo, a French bank not supported by Plaid. Although Fortuneo is available through Synci and Lunch Flow, those integrations do not import its pending or deferred card transactions. My integration already tracks their lifecycle and can reliably update or remove them once they settle, but the API currently prevents them from being represented as truly pending in Lunch Money.
J
JP Shipherd
Thanks for submitting this, Daniel. Currently, only transactions associated with Plaid-synced accounts can be pending, and Lunch Money has infrastructure in place to ensure these transactions eventually exit pending status.
If we opened it up to non-plaid transactions, I think we would need to rely on the 3rd party apps that created the original pending transactions to "clean them up", either by changing them out of pending status or deleting them if they never actually materialize. If those apps don't behave well or the connection breaks, transactions could remain pending indefinitely.
I'm curious if users would be OK with that. Perhaps if we get a few more votes on this feature request, we can find out.
D
Daniel Prilik
It seems there are some other existing asks related to being able to force-delete / toggle the pending state of plaid-based transactions (since plaid is also far from bulletproof), so if something like that were implemented, that'd certainly assuage some fears of having dangling transactions.
I guess the meta-point is that it's a bit unfortunate that there's no way to easily tell when an API transaction originate from a personal project vs. a more "production ready" 3rd party app (e.g: that might have its own support channels, or even a separate paid subscription).
I would expect most folks would be more than happy to deal with jank caused by self-hosting a 3rd party integration, whereas they'd be quite annoyed if a paid / hosted 3rd party integration ended up cluttering up their transaction feed with "zombie" pending transactions...