Dev Updates

We added offline collection payloads for Riftbound Origins so external price-index refresh work can run against the same inputs we use for other supported sets, without one-off downloads or missing-catalog errors when that set is included in a job. A small capture helper documents how those artifacts are produced so the process stays repeatable as we expand coverage. We fixed Riftbound set slug normalization so manifest identifiers line up with external product URLs (including Spiritforged-style sets), ran a bounded pilot ingest that accepted every row locally with no duplicates, and recorded overlap protection so concurrent runs for the same shard cannot clobber each other. Rollout notes capture what is validated versus what still needs timing and cleanup hardening before broad recurring use.

We strengthened the backend foundation for keeping supported sales history fresher across a wider set of releases, so expanding recurring updates for modern Pokemon and the currently supported Riftbound sets can happen more safely and predictably. That work is aimed at reducing the chance that supported cards show thinner market context simply because refresh coverage is uneven or operationally fragile. This release also tightens how a small set of legacy listing identities are resolved behind the scenes, which helps incoming sales history attach to the intended cards more reliably as coverage grows. The overall effect is better readiness for broader ongoing history refreshes without loosening the boundaries around which catalog-backed items we support.

We added a large batch of completed-sale history for the Ascended Heroes set so card and sales surfaces can reflect much richer recent market activity for that release, without widening coverage beyond the catalog-backed items we already show. Members browsing those cards should see fuller price and sales context where the index lists the card. The work completes the same carefully scoped ingest pass in production that was already validated locally, and tightens identity matching for a small set of legacy listing URL variants so rows attach to the right catalog entries. Together that keeps environments aligned and reduces gaps where history looked thin only because data had not yet landed.

Add ASC final parity closeout audit script and artifacts. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

Add ASC local off-page desupport execution script and artifacts. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

Add ASC production shared off-page desupport execution script and artifacts. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

Add ASC off-page desupport plan and production cleanup wave2 artifacts. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

Add ASC production off-page review audit script, preflight, queue CSV, and audit doc. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

Add ASC production cleanup wave1 script, preflight and metrics artifacts, and audit docs. Index-style experiences depend on consistent responses from card, set, and history surfaces; tightening this area reduces confusing gaps between what members expect and what the UI can show.

We completed the broadened Ascended Heroes production recovery under the live page-truth model after the earlier preflight proved production was missing both mappings and the underlying card rows. The pass created the exact 287 missing ASC cards on production, inserted the exact 287 approved page-truth marketplace mappings, and brought production page-truth coverage from 356 to the full 643 URLs with no off-page additions. We also resolved the one known duplicate active normalized URL by keeping the lower mapping id active and deactivating the higher duplicate row.

We ran a production reconciliation preflight for Ascended Heroes under the live page-truth model: validate the exact 287-url add list, the one duplicate active normalized URL, and exclusion sets—then emit reviewable artifacts. Preflight confirmed the add list and counts still match the repo (643 page-truth URLs, 287 missing from production, 0 local adds) but blocked execution because production has no active ASC card rows for those 287 canonical IDs, so mapping-only inserts cannot run yet. No production database writes were applied.

We documented where the 153 active Ascended Heroes mappings that sit outside the 643-row page-truth payload likely came from. Row-level evidence ties the cohort to a March 12 creation window and exact alignment with each card’s metadata marketplace URL—the pattern expected from catalog-wide bootstrap or metadata-driven backfill paths, not from the later March 28 page-truth recovery adds. The write-up maps risky code paths (bootstrap, union scrape inputs) versus safe planning-only generators, and reiterates that production’s planned add-only backfill should follow the exact page-truth add list only.