@ryuxdsgn/design-intelligence 2.18.0 → 2.19.0

package/dist/cli/index.js

@@ -463,6 +463,21 @@ const CONTEXT_FIELDS = [
                 { value: "Premium", label: "Premium" },
             ] },
     },
+    {
+        label: "Voice and tone", hint: "(voice words, how it addresses people, what it must never sound like; tone shifts by moment)",
+        ask: { flag: "voice", group: "Design", question: "How should it talk?", example: 'Hangat, lugas, pakai "kamu"', options: [
+                { value: "Warm and casual", label: "Warm and casual" },
+                { value: "Calm and professional", label: "Calm and professional" },
+                { value: "Direct and brief", label: "Direct and brief" },
+                { value: "Playful", label: "Playful", hint: "only where it costs no clarity" },
+                { value: "Reassuring", label: "Reassuring", hint: "for money, health, or waiting" },
+            ], then: { question: "What must it never sound like? Pick any.", example: "not preachy", multi: true, options: [
+                    { value: "not stiff or overly formal", label: "Stiff or overly formal" },
+                    { value: "not salesy or hyped", label: "Salesy or hyped" },
+                    { value: "not too familiar", label: "Too familiar" },
+                    { value: "no jokes when things go wrong", label: "Jokey when things go wrong" },
+                ] } },
+    },
     {
         label: "Motion direction", hint: "(feel, what motion communicates, what to avoid)",
         ask: { flag: "motion", group: "Design", question: "How much motion?", example: "Only to confirm a payment went through", options: [
@@ -768,7 +783,7 @@ Options:
   --list-rules         scan: list what each check looks for
   --ignore <ids>       scan: skip rules, e.g. RX-DS-03 (or a ryux-ignore RX-.. comment in the file)
   --widths <list>      scan <url>: viewport widths in CSS px (default 320,390,1280)
-  --product, --work, --audience, --market, --constraints, --intent, --ux, --ui, --motion <text>
+  --product, --work, --audience, --market, --constraints, --intent, --ux, --ui, --voice, --motion <text>
                        set project context fields without questions (setup)
 
 Examples:

package/dist/intelligence/analyze.js

@@ -55,8 +55,10 @@ Cover what the captures show; skip what they do not, and say it was not visible.
 | **Imagery and motion** | art direction (product UI, photography, illustration, 3D) and its style; each visual's job; composition (where the focal point sits, the space kept for copy); motion character and level (L1 to L5) when motion was captured |
 | **Design language** | the visual language in 3 to 5 traits, each tied to evidence; component patterns; the spacing and typography systems; interaction patterns |
 
-Also note content and tone (natural Indonesian or translated, how money and dates are written) and
-local patterns (QRIS, virtual accounts, OTP, addresses, paylater, e-KYC).
+Also note content and tone (natural Indonesian or translated, how money and dates are written,
+"kamu" or "Anda") and local patterns (QRIS, virtual accounts, OTP, addresses, paylater, e-KYC). When
+there is enough copy, place the voice on the four RX-CD-11 scales with the screens it comes from;
+it is observed, not the brand's rule until the user confirms it.
 
 ### 4. Patterns and evidence
 

package/dist/intelligence/content.js

@@ -661,13 +661,13 @@ export const RULES = [
             num: 11,
             title: "Tone fits the moment, voice stays",
             level: "preferred",
-            do: "Set a tone profile once, from the brand and the users, on four scales: formal to casual, serious to funny, respectful to irreverent, matter-of-fact to enthusiastic; add a few tone words and the anti-tone words to avoid (\"hangat, tidak sok akrab\").\n  - Keep the voice the same everywhere and shift the tone with the user's state: matter-of-fact and serious for errors, money, and waiting; brief for success, where a check mark already says it worked (\"Alamat disimpan\", without \"berhasil\" unless the voice wants the warmth); warmer only in onboarding and empty moments.\n  - Add personality only where it costs no clarity or length: clarity first, concision second, character third.",
+            do: "Set a tone profile once, from the brand and the users, on four scales: formal to casual, serious to funny, respectful to irreverent, matter-of-fact to enthusiastic; add a few tone words and the anti-tone words to avoid (\"hangat, tidak sok akrab\").\n  - Take it from the Voice and tone field in DESIGN.md when it is set. When it is blank, infer it from the product's existing copy or brand docs, label it inferred in Assumptions, and keep it out of DESIGN.md until the user confirms it: copy that is live is observed, not a brand rule.\n  - Keep the voice the same everywhere and shift the tone with the user's state: matter-of-fact and serious for errors, money, and waiting; brief for success, where a check mark already says it worked (\"Alamat disimpan\", without \"berhasil\" unless the voice wants the warmth); warmer only in onboarding and empty moments.\n  - Add personality only where it costs no clarity or length: clarity first, concision second, character third.",
             dont: "Joke in an error or a payment failure, cheer on a routine save, or switch between \"kamu\" and \"Anda\" across screens.",
             why: "The same words feel caring in one moment and careless in another; tone measurably changes how users judge a product.",
             basis: "NNGroup tone-of-voice dimensions (four dimensions; tested tone profiles); NNGroup 3 C's of informational microcopy",
             notWhen: "the brand guide already defines the tone; follow it",
             tradeoff: "a sober tone in serious moments gives up some brand flavour",
-            check: "review",
+            check: "review against the Voice and tone field in DESIGN.md; a profile taken from existing copy is labeled inferred",
         },
         {
             num: 12,

package/dist/intelligence/guides.js

@@ -73,17 +73,22 @@ The six stages of the core workflow (SKILL.md), as Design runs them.
    counts toward the three) instead of choosing it yourself. **No DESIGN.md and no product docs is the
    case where these questions matter most, never a reason to skip them:** ask, then wait for the
    answer before designing. Complete context means no questions.
+   **Voice and tone** is not asked by default: when the field is blank, infer it from the product's
+   existing copy or brand docs, label it inferred, and continue (RX-CD-11). Ask only when there is
+   no source at all and the work is mostly copy (onboarding, marketing, error and success messages);
+   it counts toward the three.
    **Ask as choices.** When you ask, or the user picks (stage 4, Reconsider), use the agent's
    question or choice tool if it has one: two to four options per question, the recommended one
    first and labeled, and room for the user's own answer. Without such a tool, write numbered
    options with the recommendation first, so the user can answer with a number. Never ask what the prompt, DESIGN.md, or the
    product docs already answer. For the rest, make the smallest reasonable assumption, write it
    down, and continue (RX-PR-01, RX-PR-02).
-   **Record direction answers.** When the user answers a UI or motion direction question, write the
-   answer into the UI direction or Motion direction field of the RYUX project context in DESIGN.md:
+   **Record direction answers.** When the user answers a UI direction, motion direction, or voice
+   question, or confirms an inferred voice, write the answer into the UI direction, Motion direction,
+   or Voice and tone field of the RYUX project context in DESIGN.md:
    between \`<!-- ryux-context:start -->\` and \`<!-- ryux-context:end -->\`, one line per field
    (\`- **UI direction**: Calm, trustworthy\`). If DESIGN.md or the block does not exist, create it
-   with those two fields. Change only those field lines; leave every other line as it is. Your own
+   with the fields answered. Change only those field lines; leave every other line as it is. Your own
    assumptions do not go into DESIGN.md.
    **What exists.** When there is a screen, product, or design system already, read its
    analysis or run Analyze first (capabilities/analyze.md). Existing screens, components, and any

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ryuxdsgn/design-intelligence",
-  "version": "2.18.0",
+  "version": "2.19.0",
   "description": "RYUX: design intelligence for AI agents and designers. One skill, five entry points: Analyze, Design, Build, Critique, and QA, for Claude Code, Codex, Cursor, Gemini CLI, OpenCode, Cline, Copilot, and more.",
   "type": "module",
   "bin": {