Auto-Approval Rules

HLF explicitly specified automatic approval for claims under 50 km with no additional expenses, reflecting a common pattern where the majority of submissions are routine and low-risk. Manual review of these claims consumes disproportionate coordinator time relative to the financial exposure, creating bottleneck delays that frustrate peer mentors and undermine the reimbursement system's purpose. Auto-approval rules directly reduce coordinator administrative load, accelerate reimbursement turnaround for peer mentors, and improve volunteer retention by removing friction from a core operational workflow. Organizations can calibrate thresholds to their own risk tolerance and accounting requirements, making the feature applicable across all tenant organizations without per-organization code changes. The audit trail and manual override capability ensure that automation does not compromise accountability, satisfying the compliance requirements associated with Bufdir-funded activities and accounting system integrations.

Rule configuration is stored in a dedicated table (or a JSONB configuration column on the organizations table) scoped per tenant. The auto-approval engine runs synchronously during expense claim submission: after validation passes, the engine evaluates the claim against the tenant's active ruleset and sets status to auto_approved if all conditions are met, bypassing the manual queue. Rules are expressed as a structured schema (type: distance_km | total_amount | receipt_required | expense_type_allowlist) with operator and value fields, not arbitrary code, to prevent injection and simplify the admin UI. The preview feature runs a read-only dry-run query against historical claims. The admin UI is a Next.js server component with form validation matching the backend rule schema. The Flutter app has no configuration surface; peer mentors see only the submission outcome. All rule changes are written to the audit log with before/after snapshots.