chore: trim keywords from 37 to 10, reversing the expansion

.changeset/lucky-terms-trim.md

@@ -0,0 +1,28 @@
+---
+"naija-id": patch
+---
+
+Trim `keywords` from 37 to 10, reversing the previous expansion.
+
+The expansion in 0.5.2 pushed in the wrong direction. Measuring the "nuban" result set (n=14 packages that all carry the term in their description), search score correlates **negatively** with keyword count (−0.24), description length (−0.30) and how late the term appears in the description (−0.29) — the signature of TF-IDF field-length normalisation. Every extra keyword dilutes the ones that matter. For scale, `#1 nuban` carries **2** keywords and a 47-character description; we carried 37 and 186.
+
+The expansion caused no measured harm (`nuban` was #8 before and #8 after), so this is a course correction rather than a bug fix.
+
+Kept, because these are the terms that measurably rank:
+
+| keyword | rank |
+| --- | --- |
+| `vnin` | #1 |
+| `rsa-pin` | #2 (as "rsa pin") |
+| `bvn` | #3 |
+| `naija` | #5 |
+| `nuban` | #8 |
+| `nin` | #10 |
+| `cac` | #20 |
+| `tin` | #29 |
+
+Plus `tax-id` (the current NRS identifier — high intent, no ranking data yet) and `nigeria` (the country term, and it appears in the multi-word queries).
+
+Dropped 27, all of which returned **no ranking at all** when tested directly: `kyc`, `pii`, `redaction`, `identity`, `e164`, `msisdn`, `phone`, `plate`, `passport`, `validation`, `validator` and the spelled-out variants added in 0.5.2. These sit in crowded generic fields where npm's name-token matching decides the result, so they contributed dilution and nothing else.
+
+The **description is deliberately unchanged.** Front-loading identifier names would improve term position, but it is the human-facing line on the npm page and in GitHub search results — trading readability for a couple of positions on a query where the exact-name package outscores us 13× is a bad deal.

package.json

@@ -1,45 +1,18 @@
 {
   "name": "naija-id",
   "version": "0.5.2",
-  "description": "Zero-dependency validators, formatters and PII redaction for Nigerian identifiers — phone (E.164), NIN, vNIN, BVN, Tax ID, TIN, CAC, NUBAN, plate, passport, driver's licence and RSA PIN.",
+  "description": "Zero-dependency validators, formatters and PII redaction for Nigerian identifiers \u2014 phone (E.164), NIN, vNIN, BVN, Tax ID, TIN, CAC, NUBAN, plate, passport, driver's licence and RSA PIN.",
   "keywords": [
     "nigeria",
     "naija",
-    "validation",
-    "phone",
-    "msisdn",
-    "e164",
     "nin",
-    "vnin",
     "bvn",
+    "nuban",
+    "vnin",
     "cac",
-    "rc-number",
     "tin",
     "tax-id",
-    "nuban",
-    "bank-code",
-    "plate",
-    "passport",
-    "rsa-pin",
-    "identity",
-    "kyc",
-    "pii",
-    "redaction",
-    "nigerian",
-    "validator",
-    "national-identity-number",
-    "bank-verification-number",
-    "virtual-nin",
-    "account-number",
-    "bank-account-validation",
-    "voters-card",
-    "vin",
-    "drivers-licence",
-    "fixed-line",
-    "landline",
-    "phone-validation",
-    "data-masking",
-    "log-redaction"
+    "rsa-pin"
   ],
   "license": "MIT",
   "author": "Musa Musa",