@bensandee/tooling 0.84.3 → 0.85.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@bensandee/tooling",
-  "version": "0.84.3",
+  "version": "0.85.0",
   "description": "CLI tool to bootstrap and maintain standardized TypeScript project tooling",
   "license": "MIT",
   "repository": {

package/tooling.schema.json

@@ -364,13 +364,13 @@
           "maximum": 9007199254740991
         },
         "moduleSizeLimit": {
-          "description": "Source lines a module must exceed before its importer count can make it a fan-in outlier (default 200). A module with many importers and little surface is a leaf, not a junction: it has one job and no reason to change for any of its callers, so counting its importers reports popularity rather than coupling. Applies to fan-in only -- a small module with many dependencies is a barrel, which is a finding.",
+          "description": "Raw source lines a module must exceed before its importer count can make it a fan-in outlier (default 200) -- comments and blanks included, since the question is how big the module is to read, not how much code it holds. A module with many importers and little surface is a leaf, not a junction: it has one job and no reason to change for any of its callers, so counting its importers reports popularity rather than coupling. Applies to fan-in only -- a small module with many dependencies is a barrel, which is a finding.",
           "type": "integer",
           "minimum": 0,
           "maximum": 9007199254740991
         },
         "crossPackage": {
-          "description": "Count dependency edges that cross a workspace package boundary (default true). On by default because the alternative measures the wrong thing: with it off, moving a module into its own package removes its importers from every metric, so extracting a package hides a hub rather than fixing it. A monorepo adopting the default sees fan-in and fan-out rise once, and clears it with a one-time `health:check --update-baseline --reset`; a single-package repo has no workspace siblings to resolve, so no new edge enters its graph. Set it to false only where the package boundary is a curated API surface an edge across genuinely costs less, or to hold a baseline on the old measurement basis.",
+          "description": "Count dependency edges that cross a workspace package boundary (default true). On by default because the alternative measures the wrong thing: with it off, moving a module into its own package removes its importers from every metric, so extracting a package hides a hub rather than fixing it. A monorepo adopting the default sees fan-in and fan-out rise once, and adopts that rise as a ceiling-only pull request -- a branch whose diff changes nothing but health-baseline.json and .tooling.json, which a person merges; a single-package repo has no workspace siblings to resolve, so no new edge enters its graph. Set it to false only where the package boundary is a curated API surface an edge across genuinely costs less, or to hold a baseline on the old measurement basis.",
           "type": "boolean"
         },
         "history": {