Fix IP asset fetch pipeline: owner filtering, local caching, error surfacing

[creativemuse]

Summary

• Owner filtering: GET_USER_IP_ASSETS GraphQL query now filters by caller (wallet address) so Launchpad shows only the connected user's assets
• Local DB caching: New LocalIPAsset Prisma model persists assets immediately after Crossmint registration, giving instant visibility before the Goldsky subgraph indexes the on-chain event
• Merged data sources: Launchpad fetches from both subgraph and local DB in parallel, deduplicating by storyProtocolId
• Metadata error surfacing: New metadataStatus field (loading/loaded/failed) with UI warning badge on IPCard when metadata is unavailable
• Retry with backoff: Metadata fetches from Grove/IPFS now retry up to 3 times with exponential backoff

Files changed

• lib/graphql.ts — owner filter, metadataStatus, retry logic
• prisma/schema.prisma — new LocalIPAsset model
• app/api/register-ip/route.ts — persist to local DB after Crossmint success
• app/api/my-assets/route.ts — new endpoint to query local assets by owner
• components/launchpad.tsx — merged data sources with dedup
• components/ip-card.tsx — metadata failure warning badge

Test plan

• Connect wallet and verify only your own assets appear in Launchpad
• Register a new IP asset and confirm it appears immediately (before subgraph indexes)
• Verify that once the subgraph indexes the asset, no duplicates appear
• Simulate metadata fetch failure and verify the warning badge shows
• Run npx prisma db push to apply the schema migration

🤖 Generated with Claude Code