feat(scanner): Layer-0 visual catalog search (Phase 3) (#160 )

Add pgvector embeddings on cards, server-side cohere/embed-v4.0 via AI
Gateway, kNN identify route, and L0→L1→L2 client orchestration with
empty-index fast escalate and id-cursor backfill job.

Co-authored-by: Cursor <cursoragent@cursor.com>

2026-08-14 21:46:39 -05:00

642 B

Raw Blame History

convoy

brief_number

depends_on

recommended_model

model_tier

files

scan-visual-catalog-search

composer-2.5-fast

fast

lib/card-embed.js

scripts/backfill-card-embeddings.js

package.json

test/lib/card-embed.test.js

Brief 2: Catalog embedding backfill

Goal

Server-side embed job using cohere/embed-v4.0 via AI Gateway; idempotent backfill from cards.image_url.

Acceptance criteria

lib/card-embed.js exports embedCardImage, formatEmbeddingForPg
npm run backfill-embeddings script (skips rows with embedding unless FORCE=1)
Rate-limited (SLEEP_MS) and DRY_RUN support

642 B Raw Blame History

Brief 2: Catalog embedding backfill

Goal

Acceptance criteria

642 B

Raw Blame History