Fix tokenizer: split hyphens into separate tokens

claude-memories was tokenized as one word, missing keyword matches
for entries with separate "claude" and "memories" keywords.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: mcp-tool-shop

CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 1.4.1 — 2026-03-06
+
+- **Fix tokenizer**: hyphens now split into separate tokens (`claude-memories` → `claude` + `memories`), fixing keyword matching for hyphenated task descriptions
+- Dogfooded end-to-end: install → index → resolve → planLoad against a real 32-entry workspace
+
 ## 1.4.0 — 2026-03-06
 
 - **Agent runtime contract**: `planLoad()`, `recordLoad()`, `manualLookup()` — the canonical agent integration API

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mcptoolshop/ai-loadout",
-  "version": "1.4.0",
+  "version": "1.4.1",
   "description": "Context-aware knowledge router for AI agents. Dispatch table, matcher, hierarchical resolver, agent runtime contract, CLI.",
   "type": "module",
   "bin": {

src/match.ts

@@ -21,7 +21,7 @@ function tokenize(text: string): Set<string> {
   return new Set(
     text
       .toLowerCase()
-      .replace(/[^a-z0-9\s-]/g, " ")
+      .replace(/[^a-z0-9\s]/g, " ")
       .split(/\s+/)
       .filter((w) => w.length > 1),
   );