get-shit-pretty 0.2.0 → 0.4.0

package/commands/gsp/research.md

@@ -1,6 +1,6 @@
 ---
 name: gsp:research
-description: Analyze design trends and competitive landscape for the project
+description: Deep project research — UX patterns, competitor UX, technical approaches
 allowed-tools:
   - Read
   - Write
@@ -8,64 +8,121 @@ allowed-tools:
   - Task
   - WebSearch
   - WebFetch
+  - Grep
+  - Glob
 ---
 <context>
-Phase 1 of the GSP design pipeline. Uses the Design Trend Synthesizer prompt to analyze industry trends, competitor positioning, and user expectation shifts.
+Phase 2 of the GSP project diamond. Deep research phase that investigates UX patterns, competitor experiences, technical approaches, accessibility strategies, and content patterns specific to what this project is building.
+
+This is NOT brand-level discovery (that's `/gsp:brand-discover`). This is project-level research — focused on the product type, user flows, and implementation challenges.
+
+Works with the dual-diamond architecture: reads brand context from `.design/branding/{brand}/` via `brand.ref`, reads/writes project assets in `.design/projects/{project}/`.
 </context>
 
 <objective>
-Research design trends for the project's industry and produce a comprehensive trends report.
+Deep research into UX patterns, competitor experiences, and technical approaches for this project.
 
-**Input:** `.design/BRIEF.md`
-**Output:** `.design/research/TRENDS.md`
-**Agent:** `gsp-researcher`
+**Input:** Brief scope + brand system + project BRIEF.md
+**Output:** `{project}/research/` (6 research chunks + INDEX.md)
+**Agent:** `gsp-project-researcher`
 </objective>
 
 <execution_context>
-@/Users/jubs/.claude/get-shit-pretty/prompts/07-design-trend-synthesizer.md
+@/Users/jubs/.claude/get-shit-pretty/prompts/12-project-researcher.md
 @/Users/jubs/.claude/get-shit-pretty/templates/phases/research.md
 </execution_context>
 
 <process>
+## Step 0: Resolve project and brand
+
+Scan `.design/projects/` for project directories. If only one project exists, use it. If multiple, ask the user which project to work on.
+
+Set `PROJECT_PATH` = `.design/projects/{project}`
+
+Read `{PROJECT_PATH}/brand.ref` to resolve brand path:
+- Set `BRAND_PATH` = `.design/branding/{brand}`
+
 ## Step 1: Load context
 
-Read `.design/BRIEF.md` to understand:
-- Industry / sector
-- Target audience
-- Brand personality and positioning
-- Competitive landscape (if mentioned)
+### Brief (chunk-first)
+
+Read `{PROJECT_PATH}/brief/INDEX.md`. If it exists, load `scope.md` and `target-adaptations.md`.
+
+If brief doesn't exist, tell the user to run `/gsp:brief` first.
 
-If BRIEF.md doesn't exist, tell the user to run `/gsp:new-project` first.
+### Brand system (selective)
 
-## Step 2: Spawn researcher
+Read `{BRAND_PATH}/system/INDEX.md`. If it exists, load foundation chunks (to understand the design system constraints).
 
-Spawn the `gsp-researcher` agent with:
-- The full BRIEF.md content
-- The Design Trend Synthesizer prompt (07)
+### Brand discovery (selective)
+
+Read `{BRAND_PATH}/discover/INDEX.md`. If it exists, load `competitive-audit.md` and `trend-analysis.md` (to avoid duplicating brand-level research).
+
+### Project context
+
+Read:
+- `{PROJECT_PATH}/BRIEF.md` — what we're building, platforms, audience
+- `{PROJECT_PATH}/config.json` — get `implementation_target`, `platform`, `tech_stack`
+
+## Step 1.5: Scope check
+
+**If `design_scope` is `tokens`:**
+1. Update `{PROJECT_PATH}/STATE.md` — set Phase 2 (Research) status to `skipped`
+2. Display: "Research phase skipped — design scope is `tokens`."
+3. Route: "Run `/gsp:build`."
+4. Stop here.
+
+## Step 2: Spawn project researcher
+
+Spawn the `gsp-project-researcher` agent with:
+- Brief scope chunks
+- Brand system foundation chunks
+- Brand discovery chunks (competitive audit, trends — to build on, not duplicate)
+- BRIEF.md
+- config.json preferences
+- The Project Researcher prompt (12)
 - The research output template
-- Instruction to focus on the project's specific industry
+- `implementation_target`, `platform`, `tech_stack`
+- **Output path:** `{PROJECT_PATH}/research/`
+
+The agent researches using WebSearch and writes chunks directly:
+- `research/ux-patterns.md`
+- `research/competitor-ux.md`
+- `research/technical-research.md`
+- `research/accessibility-patterns.md`
+- `research/content-strategy.md`
+- `research/reference-specs.md`
+- `research/recommendations.md`
+- `research/INDEX.md`
 
-The agent should:
-1. Research current design trends for the industry
-2. Identify 5 macro trends with real brand examples
-3. Map competitor positioning (2x2)
-4. Analyze user expectation shifts
-5. Assess platform evolution (iOS, Material, Web)
-6. Provide strategic recommendations
-7. Define mood board direction
+## Step 2.5: Write exports
 
-## Step 3: Write output
+Update `{PROJECT_PATH}/exports/INDEX.md`:
+- If INDEX.md doesn't exist, copy it from `templates/exports-index.md`
+- Replace everything between `<!-- BEGIN:research -->` and `<!-- END:research -->` with populated table:
 
-Write the completed research to `.design/research/TRENDS.md`.
+```markdown
+<!-- BEGIN:research -->
+| Section | File |
+|---------|------|
+| UX Patterns | [ux-patterns.md](../research/ux-patterns.md) |
+| Competitor UX | [competitor-ux.md](../research/competitor-ux.md) |
+| Technical Research | [technical-research.md](../research/technical-research.md) |
+| Accessibility Patterns | [accessibility-patterns.md](../research/accessibility-patterns.md) |
+| Content Strategy | [content-strategy.md](../research/content-strategy.md) |
+| Reference Specs | [reference-specs.md](../research/reference-specs.md) |
+| Recommendations | [recommendations.md](../research/recommendations.md) |
+<!-- END:research -->
+```
 
-## Step 4: Update state
+## Step 3: Update state
 
-Update `.design/STATE.md`:
-- Set Phase 1 (Research) status to `complete`
+Update `{PROJECT_PATH}/STATE.md`:
+- Set Phase 2 (Research) status to `complete`
 - Record completion date
 
-## Step 5: Route next
+## Step 4: Route next
 
-Display a summary of key findings and end with:
-"Run `/gsp:brand` to create brand identity based on these insights."
+"Run `/gsp:design` to design screens informed by this research."
 </process>
+</output>

package/commands/gsp/review.md

@@ -1,103 +1,159 @@
 ---
 name: gsp:review
-description: Design critique (Nielsen's heuristics) + WCAG accessibility audit
+description: Validate built deliverables — token compliance, screen coverage, acceptance
 allowed-tools:
   - Read
   - Write
   - Bash
   - Task
+  - Grep
+  - Glob
 ---
 <context>
-Phase 6 of the GSP design pipeline. Runs two agents in parallel: the Design Critique Partner (Nielsen's 10 heuristics) and the Accessibility Auditor (WCAG 2.2 AA). If critical issues are found, loops back to design/spec phases.
+Phase 6 of the GSP project diamond. Validates that built deliverables match the design intent — checking system token usage, screen coverage, component implementation, and accessibility compliance in the actual code.
+
+Works with the dual-diamond architecture: reads brand system from `.design/branding/{brand}/system/` via `brand.ref`, reads/writes project assets in `.design/projects/{project}/`.
 </context>
 
 <objective>
-Critique design quality and audit accessibility compliance.
+Validate built deliverables against design intent.
 
-**Input:** All prior artifacts
-**Output:** `.design/review/CRITIQUE.md` + `.design/review/ACCESSIBILITY.md`
-**Agents:** `gsp-critic` + `gsp-accessibility-auditor`
+**Input:** Built code + design chunks + brand system
+**Output:** `{project}/review/` (acceptance-report.md + issues.md + INDEX.md) + exports/INDEX.md update
+**Agent:** `gsp-reviewer`
 </objective>
 
 <execution_context>
-@/Users/jubs/.claude/get-shit-pretty/prompts/06-design-critique-partner.md
-@/Users/jubs/.claude/get-shit-pretty/prompts/08-accessibility-auditor.md
+@/Users/jubs/.claude/get-shit-pretty/prompts/11-deliverable-reviewer.md
 @/Users/jubs/.claude/get-shit-pretty/templates/phases/review.md
-@/Users/jubs/.claude/get-shit-pretty/references/nielsen-heuristics.md
-@/Users/jubs/.claude/get-shit-pretty/references/wcag-checklist.md
 </execution_context>
 
 <process>
+## Step 0: Resolve project and brand
+
+Scan `.design/projects/` for project directories. If only one project exists, use it. If multiple, ask the user which project to work on.
+
+Set `PROJECT_PATH` = `.design/projects/{project}`
+
+Read `{PROJECT_PATH}/brand.ref` to resolve brand path:
+- Set `BRAND_PATH` = `.design/branding/{brand}`
+
 ## Step 1: Load context
 
-Read `.design/config.json` to get `implementation_target`.
+Read `{PROJECT_PATH}/config.json` to get `implementation_target` and `design_scope`.
 
-Read all prior artifacts:
-- `.design/BRIEF.md`
-- `.design/brand/IDENTITY.md`
-- `.design/system/SYSTEM.md`
-- `.design/screens/SCREENS.md`
-- `.design/specs/SPECS.md` (if it exists)
+### Load all artifacts
 
-If SCREENS.md doesn't exist, tell the user to complete the design phase first.
+**Build output:** Read `{PROJECT_PATH}/build/CODE.md` and scan `{PROJECT_PATH}/build/components/` for implemented components.
 
-If SPECS.md doesn't exist and `implementation_target` is not `skip`, tell the user to run `/gsp:spec` first. When target is `skip`, review SCREENS.md + SYSTEM.md without requiring SPECS.md.
+**Design:** Read `{PROJECT_PATH}/design/INDEX.md` → load all screen chunks.
+Fallback: `{PROJECT_PATH}/screens/INDEX.md` (legacy).
 
-## Step 2: Spawn critics (parallel)
+**Brand system:** Read `{BRAND_PATH}/system/INDEX.md` → load foundation + component chunks.
+Also read `{BRAND_PATH}/system/tokens.json`.
 
-**Agent 1: gsp-critic** — Design critique using:
-- The Design Critique Partner prompt (06)
-- Nielsen's 10 Heuristics reference
-- All design artifacts
+**Brief:** Read `{PROJECT_PATH}/brief/INDEX.md` → load scope and adaptations.
 
-Should deliver:
-1. Nielsen's 10 heuristics scored 1-5 with specific examples
-2. Visual hierarchy analysis
-3. Typography and color assessment
-4. Usability evaluation
-5. Cognitive load analysis
-6. Prioritized fixes (Critical / Important / Polish)
-7. 2 alternative redesign directions
+**Research:** Read `{PROJECT_PATH}/research/INDEX.md` → load `reference-specs.md` (to verify specs were followed).
 
-**Agent 2: gsp-accessibility-auditor** — WCAG audit using:
-- The Accessibility Auditor prompt (08)
-- WCAG 2.2 AA checklist
-- All design artifacts
+**Critique:** Read `{PROJECT_PATH}/critique/INDEX.md` → load prioritized-fixes and accessibility-fixes (to verify they were addressed).
 
-Should deliver:
-1. Perceivable, Operable, Understandable, Robust checklists
-2. Mobile and cognitive accessibility checks
-3. Violations list with severity and WCAG criteria
-4. Remediation steps
-5. Accessibility statement
+**Codebase:** Scan actual codebase for implemented components (when `implementation_target` is not `figma`).
 
-## Step 3: Write output
+## Step 1.5: Scope check
 
-1. Write critique to `.design/review/CRITIQUE.md`
-2. Write audit to `.design/review/ACCESSIBILITY.md`
+**If `design_scope` is `tokens`:**
+1. Run token-audit-only review: verify token naming, scale consistency, contrast ratios, and brand alignment. Skip screen coverage and component coverage checks.
+2. Write `{PROJECT_PATH}/review/acceptance-report.md` (token-focused verdict) and `{PROJECT_PATH}/review/issues.md` (token issues only)
+3. Write `{PROJECT_PATH}/review/INDEX.md`
+4. Update `{PROJECT_PATH}/exports/INDEX.md` between `<!-- BEGIN:review -->` and `<!-- END:review -->` with populated table
+5. Update `{PROJECT_PATH}/STATE.md` — set Phase 6 (Review) to `complete` or `needs-revision`
+6. Route: display verdict and suggest `/gsp:launch` or re-run `/gsp:review`
+7. **Stop here**
 
-## Step 4: Assess results
+## Step 2: Spawn reviewer
+
+Spawn the `gsp-reviewer` agent with:
+- Build output (CODE.md + component files)
+- Design chunks
+- Brand system chunks + tokens.json
+- Brief chunks
+- Critique fixes (to verify resolution)
+- The Deliverable Reviewer prompt (11)
+- The review output template
+- **Output path:** `{PROJECT_PATH}/review/`
 
-Count critical and important issues across both reports.
+The agent writes chunks directly:
+- `review/acceptance-report.md`
+- `review/issues.md`
+- `review/INDEX.md`
 
-**If critical issues found:**
-- Display issues clearly
-- Recommend looping back: "Run `/gsp:design` or `/gsp:spec` to address critical issues, then `/gsp:review` again."
-- Update STATE.md review loop tracking
+## Step 3: Write exports
 
-**If no critical issues:**
-- Display positive summary with any polish items
-- Proceed to build
+Update `{PROJECT_PATH}/exports/INDEX.md`:
+
+```markdown
+<!-- BEGIN:review -->
+| Section | File |
+|---------|------|
+| Acceptance Report | [acceptance-report.md](../review/acceptance-report.md) |
+| Issues | [issues.md](../review/issues.md) |
+<!-- END:review -->
+```
+
+## Step 4: Assess results
+
+Read `review/acceptance-report.md` for the verdict:
+
+**Pass:** All screens implemented, tokens used correctly, accessibility compliant.
+**Conditional Pass:** Minor issues found, but shippable.
+**Fail:** Critical issues — must address before shipping.
 
 ## Step 5: Update state
 
-Update `.design/STATE.md`:
-- Set Phase 6 (Review) status to `complete` (or `needs-revision` if critical issues)
-- Record review loop count
+Update `{PROJECT_PATH}/STATE.md`:
+- Set Phase 6 (Review) status to `complete` or `needs-revision`
 - Record completion date
+- If Pass or Conditional Pass: Set Prettiness Level to 100%
+
+### Finalize git tracking
+
+1. If `git.branch` is set in config.json:
+   - Run `gh pr list --head {branch} --json url,number --limit 1` to find an open PR
+   - If found, update `git.pr` in config.json and STATE.md `## Git` table
+   - If not found, note: "No PR found for branch `{branch}`."
+   - If `gh` is not available, skip silently — leave PR field as "—"
+2. Include PR link in the CHANGELOG.md entry if available (see format below)
+
+### Update manifest + changelog
+
+1. Update `{PROJECT_PATH}/codebase/MANIFEST.md`:
+   - Update Status to `complete` (Pass) or `partial` (Conditional Pass)
+   - Populate Branch and PR lines in the manifest header from config.json `git` values
+   - Fix component paths if renamed during implementation
+
+2. Append to `.design/CHANGELOG.md`:
+   - Add entry with project name, date, brand, scope summary
+   - List added/modified components, patterns, file count
+   - Link to manifest for detail
+   - Use this format:
+     ```
+     ## [{project-name}] — {DATE}
+     > Brand: {brand} | Scope: {one-line scope from BRIEF.md}
+
+     **Added:** {component list, comma-separated}
+     **Modified:** {component list, or "—"}
+     **Patterns:** {patterns established, comma-separated, or "—"}
+     **PR:** [{#number}]({url}) or "—"
+     **Files:** {count} files touched → [manifest](./projects/{name}/codebase/MANIFEST.md)
+     ```
 
 ## Step 6: Route next
 
-If clean: "Run `/gsp:build` to translate designs to code."
-If issues: "Address the critical issues above, then run `/gsp:review` again."
+**If Pass/Conditional Pass:**
+"Project is fully pretty! All 6 project phases complete. Run `/gsp:launch` if you need marketing campaign assets, or `/gsp:progress` to see the full journey."
+
+**If Fail:**
+"Critical issues found. Address the issues in `review/issues.md`, then run `/gsp:review` again."
 </process>
+</output>

package/commands/gsp/strategy.md

@@ -0,0 +1,18 @@
+---
+name: gsp:strategy
+description: Brand strategy (alias for /gsp:brand-strategy)
+allowed-tools:
+  - Read
+---
+<process>
+Display:
+```
+🎨 GSP — Redirecting...
+
+/gsp:strategy now redirects to /gsp:brand-strategy in GSP 0.4.0.
+Brand commands now use the brand- prefix for clarity.
+```
+
+Tell the user to run `/gsp:brand-strategy` instead.
+</process>
+</output>

package/commands/gsp/system.md

@@ -1,74 +1,17 @@
 ---
 name: gsp:system
-description: Build design system foundations — color, type, spacing, components, tokens
+description: Design system (alias for /gsp:brand-patterns)
 allowed-tools:
   - Read
-  - Write
-  - Bash
-  - Task
 ---
-<context>
-Phase 3 of the GSP design pipeline. Uses the Design System Architect prompt to build a complete design system with foundations, components, and tokens.
-</context>
-
-<objective>
-Build a complete design system from the brand identity.
-
-**Input:** `.design/brand/IDENTITY.md`
-**Output:** `.design/system/SYSTEM.md` + `.design/system/tokens.json`
-**Agent:** `gsp-system-architect`
-</objective>
-
-<execution_context>
-@/Users/jubs/.claude/get-shit-pretty/prompts/01-design-system-architect.md
-@/Users/jubs/.claude/get-shit-pretty/templates/phases/system.md
-@/Users/jubs/.claude/get-shit-pretty/references/design-tokens.md
-</execution_context>
-
 <process>
-## Step 1: Load context
-
-Read:
-- `.design/brand/IDENTITY.md` — brand colors, type, personality
-- `.design/brand/palettes.json` — tints.dev generated color palettes (11-stop OKLCH scales)
-- `.design/BRIEF.md` — platforms, tech stack, constraints
-
-If IDENTITY.md doesn't exist, tell the user to run `/gsp:brand` first.
-If palettes.json exists, use its OKLCH values as the foundation for the color system tokens.
-
-## Step 2: Spawn system architect
-
-Spawn the `gsp-system-architect` agent with:
-- The IDENTITY.md and BRIEF.md content
-- The Design System Architect prompt (01)
-- The system output template
-- The design tokens reference
-
-The agent should deliver:
-1. Color system (primary, semantic, dark mode, contrast ratios, usage)
-2. Typography scale (9 levels, responsive, accessible)
-3. Grid system (12-column)
-4. Spacing system (8px base)
-5. Elevation / shadow scale
-6. Border radius tokens
-7. 30+ components with states, anatomy, usage, accessibility, code specs
-8. Design patterns
-9. Design tokens as JSON (`tokens.json`)
-10. Design principles and do's/don'ts
-
-## Step 3: Write output
-
-1. Write system to `.design/system/SYSTEM.md`
-2. Write tokens to `.design/system/tokens.json`
-
-## Step 4: Update state
-
-Update `.design/STATE.md`:
-- Set Phase 3 (System) status to `complete`
-- Record completion date
+Display:
+```
+🎨 GSP — Redirecting...
 
-## Step 5: Route next
+/gsp:system now redirects to /gsp:brand-patterns.
+Brand commands now use the brand- prefix for clarity.
+```
 
-Display system summary (foundation values, component count, token count) and end with:
-"Run `/gsp:design` to create UI/UX screens using this system."
+Tell the user to run `/gsp:brand-patterns` instead.
 </process>

package/commands/gsp/update.md

@@ -0,0 +1,102 @@
+---
+name: gsp:update
+description: Update GSP to the latest version
+---
+<objective>
+Check for GSP updates, show what's new, and run the update if the user confirms.
+</objective>
+
+<process>
+
+## Step 1 — Detect installation
+
+Check for a `VERSION` file to determine install type:
+
+1. **Local install**: `.claude/get-shit-pretty/VERSION` relative to the current working directory
+2. **Global install**: `~/.claude/get-shit-pretty/VERSION`
+
+If neither exists, tell the user GSP doesn't appear to be installed and suggest running:
+```
+npx get-shit-pretty
+```
+Then stop.
+
+Read the VERSION file to get the installed version. Store which type was detected (local or global).
+
+## Step 2 — Check latest version
+
+Run:
+```bash
+npm view get-shit-pretty version
+```
+
+If the command fails, tell the user the version check failed (they may be offline or npm is unavailable) and stop.
+
+## Step 3 — Compare versions
+
+If installed version >= latest version, tell the user:
+```
+GSP v{installed} is already up to date.
+```
+Then stop.
+
+## Step 4 — Show what's new
+
+Tell the user:
+```
+Update available: v{installed} → v{latest}
+```
+
+Fetch the changelog to show what changed:
+```bash
+curl -sf https://raw.githubusercontent.com/jubs-cloud/get-shit-pretty/main/CHANGELOG.md
+```
+
+If the fetch succeeds, extract and display the section for the latest version. If it fails, skip — changelog display is optional.
+
+## Step 5 — Warn about clean install
+
+Tell the user:
+```
+The update replaces:
+  • commands/gsp/*     (all GSP commands)
+  • get-shit-pretty/*  (runtime files)
+  • gsp-* agents       (all GSP agents)
+
+Custom files outside these prefixes are preserved.
+```
+
+## Step 6 — Confirm with user
+
+Ask the user to confirm before proceeding. If they decline, stop.
+
+## Step 7 — Execute update
+
+Based on the detected install type from Step 1:
+
+- **Local install**: run `npx get-shit-pretty@latest --claude --local`
+- **Global install**: run `npx get-shit-pretty@latest --claude`
+
+Show the output to the user.
+
+## Step 8 — Clear update cache
+
+Remove the update cache so the statusline reflects the new state:
+```bash
+rm -f ~/.claude/get-shit-pretty/.update-cache.json
+```
+
+Also check for a local cache:
+```bash
+rm -f .claude/get-shit-pretty/.update-cache.json
+```
+
+## Step 9 — Remind to restart
+
+Tell the user:
+```
+GSP updated to v{latest}.
+Restart your Claude Code session to load the new commands and agents.
+```
+
+</process>

package/commands/gsp/verbal.md

@@ -0,0 +1,18 @@
+---
+name: gsp:verbal
+description: Verbal identity (alias for /gsp:brand-verbal)
+allowed-tools:
+  - Read
+---
+<process>
+Display:
+```
+🎨 GSP — Redirecting...
+
+/gsp:verbal now redirects to /gsp:brand-verbal in GSP 0.4.0.
+Brand commands now use the brand- prefix for clarity.
+```
+
+Tell the user to run `/gsp:brand-verbal` instead.
+</process>
+</output>

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "get-shit-pretty",
-  "version": "0.2.0",
-  "description": "A design engineering system for Claude Code, OpenCode, Gemini, and Codex. Research → Brand → System → Design → Spec → Review → Build → Launch.",
+  "version": "0.4.0",
+  "description": "A design engineering system for Claude Code, OpenCode, Gemini, and Codex. Dual-diamond: Branding (discover → strategy → verbal → identity → system) + Projects (brief → research → design → critique → build → review).",
   "bin": {
     "get-shit-pretty": "bin/install.js"
   },