Docs

justin-napolitano/jayburd-bookkeeping / docs/research/automatic-account-binding-sources.md

Automatic account binding source review

Synced 2026-11-04 · Updates are made in the source repository.

Automatic account binding source review

Reviewed 2026-10-19 for issue #209. This record covers only the source sections named below. It is not a claim that every Plaid account type is supported by Jayburd or that provider metadata determines accounting treatment.

AB01 — Plaid Accounts API account type schema

Publisher and source: Plaid, Accounts API — account type schema.

Selected sections reviewed: the type, subtype, account_id, balances currency fields, account type schema, and account type/product support matrix. The page was accessed 2026-10-19.

Observed provider facts: Plaid's current schema recognizes top-level account types depository, credit, loan, investment, other and payroll. The documented depository subtypes are cash management, cd, checking, ebt, hsa, limited purpose checking, money market, paypal, prepaid and savings. Plaid publishes a separate product matrix because product availability varies by account type.

Project interpretation: Provider types and subtypes are external descriptive metadata. They may gate a reviewed Jayburd policy, but they do not establish a ledger category, normal side, ownership restriction, opening balance or financial statement presentation. Issue #209 therefore allows only depository accounts with subtype checking, savings, money market or cash management to seed a new monetary asset/debit ledger account. All other combinations fail closed and remain visible for explicit follow-up.

Access and freshness limits: This was a public web-document review. No Plaid tenant configuration, institution behavior or real account payload was inspected. Plaid can revise its schema and product matrix; a future change requires a new review and an explicit Jayburd policy change rather than automatic acceptance.

AB02 — Plaid Sandbox

Publisher and source: Plaid, Sandbox.

Selected sections reviewed: Sandbox institution, credential and multi-account test behavior. The page was accessed 2026-10-19.

Project interpretation: Sandbox supports fictional multi-account acceptance evidence. It does not prove production institution coverage, live credential handling, complete transaction history or accounting correctness.

Acceptance use

The fixed synthetic manifest for #209 must include three distinct supported depository accounts and at least one unsupported account. Tests must prove that the supported accounts receive distinct ledger/source accounts in one atomic command while the unsupported account remains visible and unbound. A provider success screen alone is not acceptance evidence.