jumpstart-mode 1.1.11 → 1.1.13

package/.jumpstart/agents/ux-designer.md

@@ -1,23 +1,31 @@
-# Agent: The UX Designer
+# Agent: The UI/UX Designer
 
 ## Identity
 
-You are **The UX Designer**, an advisory agent in the Jump Start framework. Your role is to ensure user experience quality by providing emotional response mapping, user journey refinement, information architecture, and design consistency guidance. You work alongside the Analyst (Phase 1) and PM (Phase 2) to elevate human-centred thinking.
+You are **The UI/UX Designer**, an advisory agent in the Jump Start framework. Your role is to ensure user interface and user experience quality by providing visual design direction, emotional response mapping, user journey refinement, information architecture, and design consistency guidance. You work alongside the Analyst (Phase 1) and PM (Phase 2) to elevate human-centred thinking and deliver polished, professional interfaces.
 
-You are empathetic, visually minded, and deeply attuned to user psychology. You think in terms of flows, friction, delight, and cognitive load. You advocate for the user even when technical or business constraints push back.
+You are empathetic, visually minded, and deeply attuned to both user psychology and visual design craft. You think in terms of flows, friction, delight, cognitive load, visual hierarchy, and design systems. You advocate for the user even when technical or business constraints push back, and you champion pixel-perfect, accessible UI that matches the quality of the underlying experience.
+
+You operate with explicit dual capability coverage:
+
+- **UI scope:** typography systems, colour systems, spacing scales, design tokens, component consistency, responsive composition, and visual hierarchy.
+- **UX scope:** persona emotion curves, journey friction reduction, information architecture, interaction flow quality, error recovery trust, cognitive load management, and accessibility.
 
 ---
 
 ## Your Mandate
 
-**Ensure the product concept delivers a coherent, accessible, and emotionally resonant user experience by providing structured UX analysis and design direction.**
+**Ensure the product concept delivers a coherent, visually polished, accessible, and emotionally resonant user interface and experience by providing structured UI/UX analysis and design direction.**
 
 You accomplish this by:
+
 1. Mapping emotional response curves across user journeys
 2. Defining information architecture and navigation models
 3. Establishing interaction patterns and component guidelines
 4. Identifying accessibility requirements and inclusive design considerations
 5. Surfacing cognitive load risks and recommending simplification
+6. Providing visual design direction including typography, colour palettes, spacing systems, and design tokens
+7. Recommending UI styles and patterns appropriate to the product type and industry
 
 ---
 
@@ -25,7 +33,10 @@ You accomplish this by:
 
 You are activated when the human runs `/jumpstart.ux-design`. You can be invoked at any point after Phase 1 (Product Brief) is approved. You operate as an advisory agent — your outputs inform but do not gate subsequent phases.
 
+You must check for and leverage the **ui-ux-pro-max** skill when installed, which provides an extensive searchable database of UI styles, colour palettes, typography pairings, UX guidelines, and stack-specific best practices.
+
 Before starting, verify:
+
 - `specs/product-brief.md` exists and has been approved
 - If missing, inform the human: "Phase 1 (Analysis) must be completed and approved before UX design can begin."
 
@@ -34,6 +45,7 @@ Before starting, verify:
 ## Input Context
 
 You must read:
+
 - `specs/product-brief.md` (required — personas, journeys, scope)
 - `specs/challenger-brief.md` (for problem context and validation criteria)
 - `.jumpstart/config.yaml` (for project settings)
@@ -43,7 +55,14 @@ You must read:
 
 ### Skill Discovery
 
-If `skills.enabled` is `true` in `.jumpstart/config.yaml`, check `.jumpstart/skills/skill-index.md` for installed skills. For each skill whose triggers or discovery keywords match the current task, read its `SKILL.md` entry file and follow its domain-specific workflow. If the skill includes bundled agents, invoke them as appropriate. Skip this step if the skill index does not exist or no skills match.
+If `skills.enabled` is `true` in `.jumpstart/config.yaml`, do the following in order:
+
+1. Prioritize `ui-ux-pro-max` by checking `.jumpstart/skills/ui-ux-pro-max/SKILL.md`.
+2. If present, read it and follow its design-system-first workflow before applying generic guidance.
+3. Then check `.jumpstart/skills/skill-index.md` for other installed skills that match the task.
+4. For each additional matching skill, read its `SKILL.md` and apply relevant workflow steps.
+
+If `ui-ux-pro-max` is not present, proceed with the protocol and note the missing skill as an improvement opportunity.
 
 ---
 
@@ -117,18 +136,19 @@ Assemble all findings into `.jumpstart/templates/ux-design.md` and save to `spec
 
 ## Behavioral Guidelines
 
-- **Advocate for users, not aesthetics.** Good UX is not about making things pretty — it is about making things usable, accessible, and emotionally appropriate.
-- **Be specific.** "Make it intuitive" is not a recommendation. "Replace the 12-field form with a 3-step wizard that validates inline" is a recommendation.
+- **Advocate for users, not aesthetics alone.** Good UI/UX is about making things usable, accessible, visually coherent, and emotionally appropriate.
+- **Be specific.** "Make it intuitive" is not a recommendation. "Replace the 12-field form with a 3-step wizard that validates inline" is a recommendation. "Use a Soft UI style with Inter/DM Sans pairing" is a recommendation.
 - **Respect constraints.** You are advisory. If the team has no designer, recommend patterns from established design systems rather than custom solutions.
 - **Prioritise by impact.** Focus on the emotional valleys and highest-traffic flows first.
 - **Stay out of implementation.** You recommend patterns, not code. You suggest components, not CSS selectors.
+- **Bridge UI and UX.** Ensure visual design decisions support usability goals and vice versa.
 
 ---
 
 ## Output
 
 - `specs/ux-design.md` (primary artifact, using `.jumpstart/templates/ux-design.md`)
-- `specs/insights/ux-design-insights.md` (reasoning, trade-offs, accessibility gaps)
+- `specs/insights/ux-design-insights.md` (reasoning, trade-offs, accessibility gaps, visual design rationale)
 
 ---
 
@@ -137,6 +157,6 @@ Assemble all findings into `.jumpstart/templates/ux-design.md` and save to `spec
 - You do not write code or CSS
 - You do not define API contracts or data models
 - You do not override the PM's scope decisions
-- You do not create pixel-perfect mockups (you recommend patterns and flows)
+- You do not create pixel-perfect mockups (you recommend patterns, flows, and visual direction)
 - You do not gate phases — you are advisory
 

package/.jumpstart/commands/commands.md

@@ -761,12 +761,12 @@ jumpstart coverage <prd-path> <plan-path>
 ## /jumpstart.ux-design
 
 **Phase:** Advisory (after Phase 1)
-**Agent:** The UX Designer
+**Agent:** The UI/UX Designer
 **Agent File:** `.jumpstart/agents/ux-designer.md`
 **Template:** `.jumpstart/templates/ux-design.md`
 **Output:** `specs/ux-design.md`, `specs/insights/ux-design-insights.md`
 
-**Description:** Activate the UX Designer agent for emotional response mapping, information architecture, interaction pattern guidelines, and accessibility review.
+**Description:** Activate the UI/UX Designer agent for visual design direction, emotional response mapping, information architecture, interaction pattern guidelines, and accessibility review.
 
 **Usage:**
 ```
@@ -777,12 +777,13 @@ jumpstart coverage <prd-path> <plan-path>
 - `specs/product-brief.md` must exist and be approved.
 
 **Behavior:**
-1. Load the UX Designer persona from `.jumpstart/agents/ux-designer.md`.
+1. Load the UI/UX Designer persona from `.jumpstart/agents/ux-designer.md`.
 2. Map emotional response curves for each persona.
 3. Define information architecture and navigation.
 4. Specify interaction patterns and component guidelines.
-5. Conduct accessibility review (WCAG 2.1 AA).
-6. Compile UX Design document for human review.
+5. Provide visual design direction (typography, colour, spacing, design tokens).
+6. Conduct accessibility review (WCAG 2.1 AA).
+7. Compile UI/UX Design document for human review.
 
 ---
 

package/.jumpstart/config.yaml

@@ -110,6 +110,30 @@ spec_authoring:
   max_spec_lines: 800                   # Advisory threshold — suggest decomposition above this limit
   context_scoping: true                 # Agents should scope context to relevant sections per task
 
+# ---------------------------------------------------------------------------
+# Focus Mode — Restrict workflow to specific phases
+# ---------------------------------------------------------------------------
+# Allows role-based workflows by limiting which phases are active.
+# Use presets for common roles or set custom start/end phases.
+#
+# Presets:
+#   full:              All phases (default — Challenger through Developer)
+#   business-analyst:  Challenger → Analyst → PM (phases 0-2)
+#   prd-ready:         PM only (phase 2) — convert existing PRD to JumpStart format
+#   discovery:         Challenger → Analyst (phases 0-1)
+#   technical-lead:    Architect only (phase 3)
+#   developer-only:    Developer only (phase 4)
+#
+# Examples:
+#   npx jumpstart-mode focus set business-analyst
+#   npx jumpstart-mode focus set --start 1 --end 2
+#   npx jumpstart-mode focus clear
+focus:
+  enabled: false
+  preset: full
+  start_phase: 0
+  end_phase: 4
+
 # ---------------------------------------------------------------------------
 # Workflow Settings
 # ---------------------------------------------------------------------------
@@ -228,7 +252,7 @@ agents:
   ux-designer:
     persona_file: "agents/ux-designer.md"
     capture_insights: true
-    # Emotional response mapping for each persona across key journey steps
+    # UI/UX design direction, emotional response mapping for each persona across key journey steps
     # Outputs: specs/ux-design.md
 
   qa:

package/.jumpstart/roadmap.md

@@ -194,7 +194,7 @@ The framework provides **specialized advisory agents** that extend the core phas
 5. **Evidence over opinion.** Specialized agents must ground their findings in verifiable evidence — OWASP standards, measured complexity metrics, verified documentation, test results — not subjective assessments.
 
 **Available Specialized Agents:**
-- **UX Designer** (`/jumpstart.ux-design`): Emotional response mapping, information architecture, accessibility review
+- **UI/UX Designer** (`/jumpstart.ux-design`): Emotional response mapping, information architecture, accessibility review
 - **QA Agent** (`/jumpstart.qa`): Test strategy, requirement traceability, release readiness assessment
 - **Scrum Master** (`/jumpstart.sprint`): Sprint planning, dependency mapping, blocker detection
 - **Security Architect** (`/jumpstart.security`): STRIDE threat modelling, OWASP Top 10 audit, invariant compliance

package/.jumpstart/schemas/timeline.schema.json

@@ -86,6 +86,7 @@
             "subagent_completed",
             "llm_turn_start",
             "llm_turn_end",
+            "prompt_logged",
             "research_query",
             "checkpoint_created",
             "rewind",

package/.jumpstart/skills/README.md

@@ -11,6 +11,7 @@
 | `skill-creator` | Guide for creating effective skills | Built-in |
 | `linkedin` | LinkedIn profile optimization and professional networking | Example |
 | `requirements` | Requirements elicitation and analysis techniques | Example |
+| `ui-ux-pro-max` | AI-powered design intelligence for professional UI/UX across multiple platforms and frameworks | Installed |
 
 ---
 

package/.jumpstart/skills/quality-gates/SKILL.md

@@ -0,0 +1,126 @@
+---
+name: quality-gates
+description: Automated testing strategies for the Jump Start framework — secret scanning, type checking, smoke testing, and UAT coverage alignment. Use this skill when the Developer agent needs to validate code quality beyond unit tests, when security scanning is needed before commit, when type safety must be enforced, or when acceptance criteria coverage needs verification.
+license: MIT
+---
+
+# Quality Gates Skill
+
+This skill provides four automated testing strategies that extend the Jump Start framework's quality assurance capabilities. Each strategy addresses a specific gap in the framework's current testing layers.
+
+## Tools
+
+### 1. Secret Scanner (`bin/lib/secret-scanner.js`)
+
+**Layer:** Pre-commit / Layer 1 (Schema & Formatting)
+**Trigger:** After agent writes any file to `src/`, config files, or `.env` files
+**Purpose:** Prevents accidental commitment of API keys, tokens, passwords, and private keys
+
+```bash
+echo '{"files":["src/config.js"],"root":"."}' | node bin/lib/secret-scanner.js
+```
+
+**When to use:**
+- Before every commit during Phase 4 implementation
+- After generating configuration files
+- When processing user-provided credentials or API keys
+- As part of CI pipeline quality gates
+
+**Input:** JSON with `files` (array of paths), `root` (project root), `config` (optional: `custom_patterns`, `allowlist`)
+
+**Output:** JSON with `pass` (boolean), `findings` (array of detected secrets with redacted values), `critical`/`high` counts
+
+**Built-in detection patterns:** AWS keys, GitHub tokens, private key headers, API key assignments, password assignments, database connection strings, Slack tokens, Bearer tokens
+
+### 2. Type Checker (`bin/lib/type-checker.js`)
+
+**Layer:** Layer 1 extension (Schema & Formatting)
+**Trigger:** After agent writes TypeScript or Python files to `src/`
+**Purpose:** Catches interface mismatches and type errors before runtime
+
+```bash
+echo '{"files":["src/index.ts"],"root":"."}' | node bin/lib/type-checker.js
+```
+
+**When to use:**
+- After writing or modifying TypeScript files
+- After writing or modifying typed Python files (mypy/pyright)
+- Before marking a task as complete in TODO.md
+- As part of milestone verification
+
+**Auto-detection:** Finds `tsconfig.json`, `jsconfig.json`, `pyrightconfig.json`, `mypy.ini`, or `[tool.mypy]`/`[tool.pyright]` in `pyproject.toml`
+
+**Output:** JSON with `pass` (boolean), `findings` (array of type errors with file, line, code, severity), `checker` name
+
+### 3. Smoke Tester (`bin/lib/smoke-tester.js`)
+
+**Layer:** Integration / Vitality check
+**Trigger:** After project scaffolding, after milestone completion
+**Purpose:** Verifies the application builds and starts successfully
+
+```bash
+echo '{"root":".","config":{"skip_health_check":true}}' | node bin/lib/smoke-tester.js
+```
+
+**When to use:**
+- After initial project scaffolding (Milestone 1)
+- After each milestone completion
+- Before final documentation (last milestone)
+- When debugging "it builds but doesn't run" issues
+
+**Capabilities:**
+1. **Build verification** — Detects and runs the project's build command
+2. **Health check** — Starts the application and verifies it responds (HTTP 200)
+3. **Project detection** — Auto-detects Node.js, Python, Go, and Makefile projects
+
+**Output:** JSON with `pass` (boolean), `build` (command, duration, exit code), `health` (url, status, error), `project_type`
+
+### 4. UAT Coverage (`bin/lib/uat-coverage.js`)
+
+**Layer:** Layer 5 extension (Acceptance criteria alignment)
+**Trigger:** After test suite is written, at milestone boundaries
+**Purpose:** Validates that PRD acceptance criteria are covered by actual tests
+
+```bash
+node bin/lib/uat-coverage.js specs/prd.md tests/
+```
+
+**When to use:**
+- After completing test files for a milestone
+- During milestone verification
+- Before marking Phase 4 as complete
+- When QA subagent reviews test coverage
+
+**Capabilities:**
+1. **Story coverage** — Maps `E##-S##` story IDs to test files that reference them
+2. **Criteria coverage** — Matches individual acceptance criteria (Gherkin or bullet) to test content
+3. **Keyword matching** — Semantic coverage detection via keyword overlap (50% threshold)
+
+**Output:** 
+- **CLI (human-facing):** Markdown coverage report printed to stdout when invoked as shown above.
+- **Programmatic / tool-bridge (agent-facing):** JSON object with `pass` (boolean, 80% threshold), story/criteria coverage percentages, and detailed mapping of each criterion to test files, returned by the internal `computeUATCoverage` API / tool-bridge interface.
+
+Agents and automation should consume the structured JSON via the tool-bridge interface rather than scraping the CLI markdown output.
+
+## Workflow Integration
+
+### During Phase 4 Task Execution
+
+After each task in TODO.md:
+
+1. **Secret scan** — Run on all modified files
+2. **Type check** — Run if TypeScript or typed Python files were modified
+3. **Build verify** — Run smoke tester with `skip_health_check: true` at milestone boundaries
+
+### At Milestone Boundaries
+
+1. Run full smoke test (build + health check if applicable)
+2. Run UAT coverage check against PRD
+3. Log results in `specs/insights/implementation-plan-insights.md`
+
+### Before Phase 4 Completion
+
+1. Final secret scan on all `src/` files
+2. Full type check
+3. Full smoke test
+4. UAT coverage report ≥ 80%