Great job putting this together Aana ! >> Let's leave this for now
+1 for this Aman. The question I would ask is "Is there anybody wanting to deploy what is being proposed in production ?" Assuming the answer is no, I would not spend any effort or accumulate tech debt for code we have to maintain because we are building it for reasons like feature parity or exposing something which exists in the core. Questions I would ask include "Is there anybody wanting to even use the self-service app for things like Savings or Origination the way it is being built right now" ? My suggestion would be to focus all available effort on the excellent list for cleaning up the core itself which was circulated by Adam / Aleks and progressive loans and making the core smaller / modular / more performant. If there aren't enough tasks on the core, then spend effort the back-office as it is useful for understanding the scope of our offerings. Given how easy it is to build new things right now, we should avoid the temptation to build for no reason other than we can. Regards, Vishwas On Fri, 13 Nov 2026 at 05:35, Aman Mittal <[email protected]> wrote: > Hi Anna +1 for your efforts > > I had to save this email from my spam folder. > > I will try to answer as best as possible. > > For > 1.Is wrapping the Core share-account API the right approach, consistent > with savings and loans? > Answer : Yes, because changes are evolving. And wrapping core APIs help us > catch and refractor any changes done in the core fineract in the PR itself. > Similar thing is also done with backoffice Ui (any breaking api changes are > caught early). So we don't need to handle lot of things. > > 2. For the consumer command allow-list, withdraw own pending application, > apply additional shares, is that the right set, or should the first version > be read-only until the apply flow is agreed? > > Ans : I will let others answer this. If no one answers (72 hours) propose > your understanding. Before implementing it. > > > 3. Should share dividends be surfaced to consumers at all, or left > entirely back-office? > Let's leave this for now : not have any opinions for it in terms of > business process. I will let community answer this. > > Regards > Aman > > > > > On Fri, 13 Nov, 2026, 5:58 pm Aana Pandey, <[email protected]> wrote: > >> Hi all, >> >> I’d like to add share accounts to fineract-consumer-facing ( >> https://github.com/apache/fineract-consumer-facing), one of the >> self-service parity gaps in #144. The docs record the gap: summary.adoc >> notes the app has “no cross-account totals, share accounts, or currency >> grouping,” and there’s a feature doc for every account type the app does >> surface - savings, loans, transfers - but none for shares, because a member >> currently can’t view their share accounts, see share products, or apply for >> shares. The removed self-service API provided this under >> /self/shareaccounts and /self/products/share. >> >> Proposal: wrap Fineract Core’s existing share-account and product APIs >> behind the BFF, with a minimal Angular UI. No Core changes are needed, the >> operations are already in the vendored spec under the Share Account tag >> (/v1/accounts/{type}, type share) and the Products tag >> (/v1/products/{type}), so this only needs those clients generated and the >> usual thin adapter, the same shape as savings and loans. >> >> A customer could: >> >> - view their share accounts and the share products on offer; >> - apply for shares, and modify or withdraw a pending application; >> - (later) apply for additional shares on an active account. >> >> Approve, activate and reject stay out of scope as back-office actions. >> Fineract multiplexes them with consumer actions on POST >> /accounts/{type}/{accountId}, so the BFF would allow-list only the >> consumer-safe commands and reject the rest, a consumer can never approve >> their own application. This follows the same deliberate narrowing the app >> already applies elsewhere (for example, user.adoc explains why consumer >> image upload is excluded even though Core allows it). >> >> Rules, reusing existing patterns: >> >> - submit/modify/withdraw go through the initiate/confirm OTP step-up >> with an Idempotency-Key, consistent with the savings-application work in >> #155; >> - reads authorize a share-scoped ABAC action per account, as the >> other account features do. >> >> Slices, one small PR each behind a flag that’s off by default: the >> Fineract adapter, then list and detail plus the products catalog >> (read-only), then submit/modify/withdraw, then the Angular screens with >> Cucumber and Playwright tests. >> >> Questions: >> >> 1. Is wrapping the Core share-account API the right approach, >> consistent with savings and loans? >> 2. For the consumer command allow-list, withdraw own pending >> application, apply additional shares, is that the right set, or should the >> first version be read-only until the apply flow is agreed? >> 3. Should share dividends be surfaced to consumers at all, or left >> entirely back-office? >> >> I’ve been contributing across loans, beneficiaries and savings (several >> merged PRs), and I’d like to take this on. Happy to adjust scope based on >> feedback before opening a tracking issue. >> >> Thanks, >> Aana >> >
