auto-bean-categorize

Categorize a parsed bank statement by assigning transaction categories and determining reconciliation or deduplication actions. This process handles statement-scoped categorization and produces a categorized artifact for downstream Beancount posting.

Statement reconciliation identifies duplicate transactions and flags pending questions within a single parsed artifact. It enforces Beancount-safe checks and generates memory suggestions to ensure accurate downstream posting without balance errors.

No, the categorization process operates on exactly one assigned parsed or intermediate artifact per activation. This ensures deterministic focus and governance, preventing cross-statement data contamination during the reconciliation process.

Handle deduplication by evaluating a single parsed-evidence artifact to determine if transactions match existing records. The workflow identifies duplicates and outputs deduplication actions within a categorized artifact for review before final posting.

Yes, you need a ready_for_categorization artifact provided by an import workflow. The categorization process requires this assigned parsed or intermediate artifact to evaluate transactions and produce accurate reconciliation findings and posting inputs.

The process is limited to a single artifact per activation and cannot perform cross-statement reconciliation. It focuses strictly on statement-scoped categorization, leaving final writing actions to the broader import workflow.