replen 1.0.30 → 1.0.32

package/extras/skills/replen/SKILL.md

@@ -51,6 +51,14 @@ Parse the JSON response. Note:
   `goal` is worth surfacing even if it doesn't map to an existing capability.
   Lead your writeups with mission-fit, not just tech-fit. (Null until the project
   is onboarded with a thesis — then fall back to capability-fit alone.)
+- `keystoneUpgrades` — task-scoped `better_than` upgrades for solutions this
+  project already uses (from Keystone, the comparative-knowledge ontology): each
+  is `{current, better, betterKind, task, margin, source}`. A deprecated/inferior
+  dependency has a maintained/stronger replacement. The footnote may surface one;
+  if the user wants it, DO the swap (update the dep, adjust call sites). Always
+  cite the `task` — "better" is task-relative (e.g. text-embedding-3-large beats
+  3-small for general retrieval but NOT for short-label matching). Don't push an
+  upgrade the user didn't ask about beyond the one calm line.
 - `candidates[]` — the actual list to triage
 - `candidates[].priorContext` — server-attached MEMORY: the user's earlier
   verdicts on this repo, and whether the matched capability is already

package/extras/skills/replen-onboard/SKILL.md

@@ -255,6 +255,19 @@ strings. Each is `{tag, descriptor, modality}`:
 Break broad capabilities into the concrete techniques the code uses. Be
 specific — `cloudflare bypass`/`proxy rotation`, not just `web scraping`.
 
+**Also capture architectural PRACTICES as capabilities.** Beyond what the code
+*does*, note the distinctive structural *moves* it makes — these are what let
+Replen suggest a pattern to your OTHER projects ("Aegis made its domain model
+data-driven; this data-heavy project should consider it"). Include a capability
+for any deliberate practice you find: a **data-driven domain ontology** (entity
+types/relationships stored as DATA/rows, not hardcoded — configurable without a
+redeploy), **event sourcing**, **outbox pattern**, **CQRS**, **pgvector
+semantic search**, **feature store**, etc. Use the standard practice name as the
+`tag` (so it matches Keystone's practice registry). Respect the cover — describe
+the practice's STRUCTURE, never the sensitive entity names it models (for a
+covered repo, "a configurable Postgres-backed domain ontology — entity types,
+typed properties, computed properties as rows", never the entity names).
+
 ### 2e. Push to Replen
 
 1. **Register** the repo if it isn't already. The per-repo `replen_set_tags` /

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "replen",
-  "version": "1.0.30",
+  "version": "1.0.32",
   "description": "Make your AI coding tools smarter. One command, no API keys, free. Replen watches what your projects actually do and surfaces a few things worth bringing in each month. Use one as is, port a piece of another, cherry pick an idea, or build it clean room. The match happens inside your AI tool's session. A few actionable matches a month, by design.",
   "type": "module",
   "bin": {