REPOSITORY CODING · HARD
Spokeflow: repair settlement and saved analytics
A bike transfer has exactly one source and one destination. The source needs enough stock and the destination enough empty docks; cargo cycles have their own limit and electric support needs approval. Cargo transfers avoid the truck startup fee. Dashboard imbalance reduction compares both endpoints before and after each confirmed transfer. The settlement calculation and frontend metric both violate the contract. Preserve rounding, unit conversion, cancellation and repeated-plan behavior while repairing them. The repository includes a resource specification inspector, native database, booking timeline and server-derived insights. Run bash test/run.sh. Browser checks include mobile confirmation, durable reload, independent HTTP probes and regression safeguards.
Work through a real codebase.
Choose an environment, investigate the existing implementation and use the tests to check your change. Sign in to open a workspace; execution and Ono usage follow your plan’s allowances.
Task requirements and environments
MEAN · 80 minute session
- A bike transfer has exactly one source and one destination. The source needs enough stock and the destination enough empty docks; cargo cycles have their own limit and electric support needs approval. Cargo transfers avoid the truck startup fee. Dashboard imbalance reduction compares both endpoints before and after each confirmed transfer.
- Use real React (PERN/MERN) or Angular (MEAN) resource cards, specification disclosure, typed planning controls, authoritative previews, confirmation dialogs, durable activity and responsive navigation. The booking timeline and server-derived insights use saved data. No product flow requires JSON editing.
- Tenant-scoped PostgreSQL row-lock transactions and MongoDB compare-and-swap aggregates persist resources, bookings, receipts and audit records. Failed journals roll back all state, repeated request IDs replay results, and stale application revisions return 409. Capacity counts the maximum simultaneous confirmed usage across half-open intervals.
- All data is synthetic and bounded. Examples use integer planning coordinates; the per-product policy defines their units. No external API, payment, trade, claim or physical equipment action is performed. Run bash start.sh for preview, npm run build for the frontend bundle, or bash test/run.sh for real browser acceptance checks.
MERN · 80 minute session
- A bike transfer has exactly one source and one destination. The source needs enough stock and the destination enough empty docks; cargo cycles have their own limit and electric support needs approval. Cargo transfers avoid the truck startup fee. Dashboard imbalance reduction compares both endpoints before and after each confirmed transfer.
- Use real React (PERN/MERN) or Angular (MEAN) resource cards, specification disclosure, typed planning controls, authoritative previews, confirmation dialogs, durable activity and responsive navigation. The booking timeline and server-derived insights use saved data. No product flow requires JSON editing.
- Tenant-scoped PostgreSQL row-lock transactions and MongoDB compare-and-swap aggregates persist resources, bookings, receipts and audit records. Failed journals roll back all state, repeated request IDs replay results, and stale application revisions return 409. Capacity counts the maximum simultaneous confirmed usage across half-open intervals.