command-code 0.31.1 → 0.32.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "command-code",
-  "version": "0.31.1",
+  "version": "0.32.0",
   "description": "Command Code, coding agent that continuously learns your coding taste",
   "type": "module",
   "main": "dist/index.mjs",

package/skills/design/SKILL.md

@@ -17,6 +17,18 @@ When the *style* is ambiguous, decide. When the *goal* is ambiguous, ask only if
 
 Do not ask for confirmation before acting on a complete brief. Infer ordinary details, choose the strongest interpretation, and ship.
 
+## Explicit report modes: smell, checkup, review
+
+**CRITICAL RULE: When the user explicitly runs `/design smell`, `/design checkup`, or `/design review`, I ONLY generate the report. I never apply fixes in the same turn.**
+
+These modes report findings only. Fixes happen when the user runs `/design redesign`, `/design relayout`, `/design recolor`, etc. — a separate, explicit command.
+
+- `/design smell` → generates `smell-report.md` and `smell-report.html` only
+- `/design checkup` → generates `checkup-report.md` and `checkup-report.html` only
+- `/design review` → generates `review-report.md` and `review-report.html` only
+
+No design changes. No fixing. No calling other modes. Just reports.
+
 ## Bare `/design` routing
 
 When the user runs `/design` with no tool and no freeform prompt, I do not show the table. I route immediately and act.

package/skills/design/references/smell.md

@@ -129,7 +129,9 @@ I name the dominant smell. I identify the root reflex. Then I pick the right des
 - Missing state and interaction smell goes to interaction
 - Unclear copy goes to writing
 
-If the design smells generated in several systems at once, redesign is usually cleaner than incremental repair.
+If the design smells in several systems at once, I suggest redesign is usually cleaner than incremental repair.
+
+**Important:** Smell never executes any mode. It only produces `.commandcode/design/smell-report.md` and `.commandcode/design/smell-report.html`.
 
 ---