[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C0HWK6XZ8RVZRMSZHTP175":3,"skill-same:01M7C0HWK6XZ8RVZRMSZHTP175":44},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M7C0HWK6XZ8RVZRMSZHTP175","visual-exploration",null,"Layer 1 skill for browser automation and visual UX documentation. Initial reconnaissance, navigation mapping, interaction documentation, end-to-end flow capture, edge case exploration, screenshot methodology. Loaded by the analyzer agent during Layer 1.","zak-keown\u002Fmoe","zak-keown","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F320717241?v=4","01M7BZFXME0NW6TQX0H7AZM2KS","01M7BZFXN32VR0JN5AATRD53H4",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe-backstory\u002F.opencode\u002Fskills\u002Fvisual-exploration\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: visual-exploration\ndescription: Layer 1 skill for browser automation and visual UX documentation. Initial reconnaissance, navigation mapping, interaction documentation, end-to-end flow capture, edge case exploration, screenshot methodology. Loaded by the analyzer agent during Layer 1.\n---\n\n# Visual Exploration Methodology\n\nDocument behavioral intelligence through browser automation and visual observation. Every screenshot is an empirical observation. Every interaction sequence is a behavioral flow. This mode captures what the user SEES and what the system DOES in response to user actions.\n\n## When to Use This Mode\n\nVisual exploration activates when:\n- The target is a web application with a UI\n- The discovery inventory identifies a running web UI accessible via HTTP\n- A running instance of the target is accessible via HTTP\n\nThis mode requires a running instance of the target application and browser automation capabilities (Playwright, Puppeteer, or equivalent). All output is **RAW** (screenshots and flow documentation capture the target's UI in detail).\n\nTargets without a web UI get no useful signal from this mode — skip it.\n\n## Why Visual Exploration Matters\n\nSource code tells you what the system CAN do. Tests tell you what the system MUST do. Visual exploration tells you what the system LOOKS LIKE while doing it. This is behavioral intelligence that no other mode captures:\n\n- **Affordances** -- what actions does the UI present as available?\n- **Feedback** -- how does the system communicate success, failure, and state changes?\n- **Navigation structure** -- how are features organized from the user's perspective?\n- **State representation** -- how does the UI change as the system's state changes?\n- **Accessibility** -- is the interface usable with keyboard, screen reader, or at different sizes?\n\n## Phase 1: Initial Reconnaissance\n\n**Goal:** Navigate to the root URL, capture the landing page, and identify the application type, auth requirements, and available affordances.\n\n### 1.1 Landing Page Capture\n\n```bash\n# Navigate to root URL and screenshot\n# (Using Playwright as reference; adapt to available browser automation)\n```\n\n```javascript\nconst page = await browser.newPage();\nawait page.goto(ROOT_URL, { waitUntil: 'networkidle' });\nawait page.screenshot({ path: 'workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fscreenshots\u002F001-landing.png', fullPage: true });\n```\n\n### 1.2 Initial Assessment\n\nDocument:\n- **Application type** -- SPA, MPA, dashboard, wizard, documentation site, etc.\n- **Authentication** -- does the landing page require login? Is there a signup flow?\n- **Primary navigation** -- top nav, sidebar, hamburger menu, tabs, breadcrumbs?\n- **Visible affordances** -- buttons, links, forms, search bars, dropdowns visible on the landing page\n- **Branding and layout** -- header, footer, sidebar, content area dimensions\n\n### 1.3 Auth Handling\n\nIf authentication is required:\n1. Screenshot the login page\n2. Document the auth method (username\u002Fpassword, OAuth, SSO, API key)\n3. If test credentials are available, authenticate and screenshot the post-login state\n4. If no credentials are available, document the gate and explore only the unauthenticated surface\n\n## Phase 2: Navigation Mapping\n\n**Goal:** Click every navigation element, screenshot each page, and document the complete navigation tree.\n\n### 2.1 Identify Navigation Elements\n\n```javascript\n\u002F\u002F Extract all navigation links\nconst navLinks = await page.$$eval('nav a, [role=\"navigation\"] a, .sidebar a, .menu a, header a',\n  links => links.map(a => ({ text: a.textContent.trim(), href: a.href }))\n);\n```\n\n### 2.2 Systematic Traversal\n\nFor each navigation element:\n1. Click the element (or navigate to the href)\n2. Wait for the page to settle (network idle or DOM stable)\n3. Take a full-page screenshot\n4. Record the URL, page title, and visible heading\n5. Identify sub-navigation elements on the new page\n\n```javascript\nfor (const [index, link] of navLinks.entries()) {\n  await page.goto(link.href, { waitUntil: 'networkidle' });\n  await page.screenshot({\n    path: `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fscreenshots\u002Fnav-${String(index).padStart(3, '0')}-${slugify(link.text)}.png`,\n    fullPage: true\n  });\n}\n```\n\n### 2.3 Navigation Tree Output\n\nWrite to `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fnavigation-map.md`:\n\n```markdown\n## Navigation Tree\n\n- **Home** (`\u002F`) -- screenshot: 001-landing.png\n  - **Dashboard** (`\u002Fdashboard`) -- screenshot: nav-001-dashboard.png\n    - **Analytics** (`\u002Fdashboard\u002Fanalytics`) -- screenshot: nav-002-analytics.png\n    - **Reports** (`\u002Fdashboard\u002Freports`) -- screenshot: nav-003-reports.png\n  - **Settings** (`\u002Fsettings`) -- screenshot: nav-004-settings.png\n    - **Profile** (`\u002Fsettings\u002Fprofile`) -- screenshot: nav-005-profile.png\n    - **Billing** (`\u002Fsettings\u002Fbilling`) -- screenshot: nav-006-billing.png\n```\n\n## Phase 3: Interaction Documentation\n\n**Goal:** For each page, identify interactive elements, interact with them, and capture before\u002Fafter screenshots documenting the system's feedback.\n\n### 3.1 Element Inventory\n\nFor each page, identify:\n- **Forms** -- text inputs, selects, checkboxes, radio buttons, textareas, file uploads\n- **Buttons** -- submit buttons, action buttons, toggle buttons\n- **Dropdowns** -- select menus, custom dropdown components\n- **Modals\u002Fdialogs** -- elements that trigger overlay content\n- **Expandable sections** -- accordions, collapsible panels, \"show more\" links\n- **Interactive widgets** -- date pickers, sliders, color pickers, drag-and-drop zones\n\n```javascript\nconst interactiveElements = await page.$$eval(\n  'button, input, select, textarea, [role=\"button\"], [role=\"tab\"], [onclick], a[href=\"#\"]',\n  els => els.map(el => ({\n    tag: el.tagName,\n    type: el.type || '',\n    text: el.textContent?.trim().substring(0, 50),\n    id: el.id,\n    name: el.name,\n    ariaLabel: el.getAttribute('aria-label'),\n  }))\n);\n```\n\n### 3.2 Interaction Capture\n\nFor each interactive element:\n\n1. **Before screenshot** -- capture the page state before interaction\n2. **Interact** -- click, type, select, toggle\n3. **After screenshot** -- capture the page state after interaction\n4. **Document feedback** -- what changed? Error message? Success toast? New content?\n\n```javascript\n\u002F\u002F Example: form submission\nawait page.screenshot({ path: `screenshots\u002Fform-${name}-before.png` });\nawait page.fill('input[name=\"email\"]', 'test@example.com');\nawait page.click('button[type=\"submit\"]');\nawait page.waitForTimeout(1000); \u002F\u002F Allow time for feedback\nawait page.screenshot({ path: `screenshots\u002Fform-${name}-after.png` });\n```\n\n### 3.3 Page Inventory Output\n\nWrite to `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fpage-inventory.md`:\n\n```markdown\n## Page: Settings > Profile\n\n**URL:** `\u002Fsettings\u002Fprofile`\n**Screenshot:** nav-005-profile.png\n\n### Interactive Elements\n\n| Element | Type | Action Taken | Feedback | Screenshots |\n|---------|------|-------------|----------|-------------|\n| Display Name | text input | Typed \"Test User\" | No immediate feedback | form-profile-before.png, form-profile-after.png |\n| Save | submit button | Clicked | Success toast: \"Profile updated\" | form-profile-save.png |\n| Avatar | file upload | Uploaded test.png | Preview updated | form-profile-avatar.png |\n```\n\n## Phase 4: Flow Documentation\n\n**Goal:** Document end-to-end user flows with screenshots at every state transition.\n\n### 4.1 Core Flows to Document\n\nDocument these flows whenever they exist:\n\n| Flow | Description | Priority |\n|------|-------------|----------|\n| **Onboarding** | First-time user experience from signup to first use | High |\n| **Core workflow** | The primary task the application enables | High |\n| **Error recovery** | What happens when something goes wrong | High |\n| **Settings\u002Fconfiguration** | How users customize the application | Medium |\n| **Search\u002Ffilter** | How users find content within the application | Medium |\n| **CRUD operations** | Create, read, update, delete for primary entities | Medium |\n| **Authentication flow** | Login, logout, password reset, session expiry | Medium |\n| **Export\u002Fimport** | How data moves in and out of the application | Lower |\n\n### 4.2 Flow Documentation Format\n\nFor each flow, create a directory under `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fflows\u002F`:\n\n```\nworkspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fflows\u002F\n    onboarding\u002F\n        step-01-landing.png\n        step-02-signup-form.png\n        step-03-email-verification.png\n        step-04-profile-setup.png\n        step-05-first-dashboard.png\n        flow.md\n    core-workflow\u002F\n        step-01-start.png\n        step-02-input.png\n        step-03-processing.png\n        step-04-result.png\n        flow.md\n```\n\nEach `flow.md` follows this format:\n\n```markdown\n# Flow: Onboarding\n\n## Steps\n\n### Step 1: Landing Page\n**Screenshot:** step-01-landing.png\n**URL:** `\u002F`\n**Action:** Click \"Sign Up\" button\n**State:** Unauthenticated, no account\n\n### Step 2: Signup Form\n**Screenshot:** step-02-signup-form.png\n**URL:** `\u002Fsignup`\n**Action:** Fill email, password, click \"Create Account\"\n**State:** Unauthenticated, form displayed\n**Feedback:** Form validation on blur for email format\n\n### Step 3: Email Verification\n**Screenshot:** step-03-email-verification.png\n**URL:** `\u002Fverify`\n**Action:** (requires email access -- documented but not tested)\n**State:** Account created, not verified\n**Feedback:** \"Check your email\" message displayed\n\n## Behavioral Claims\n\n- New accounts require email verification before first use\n  \u003C!-- cite: source=visual-observation, ref=flows\u002Fonboarding\u002Fstep-03-email-verification.png, confidence=confirmed, agent=visual-explorer -->\n```\n\n## Phase 5: Edge Cases\n\n**Goal:** Explore edge cases that reveal the system's behavior at its boundaries.\n\n### 5.1 Empty States\n\n```javascript\n\u002F\u002F Navigate to pages likely to have empty states (lists, dashboards, search results)\n\u002F\u002F Screenshot each empty state\nawait page.goto('\u002Fprojects'); \u002F\u002F Assuming no projects exist\nawait page.screenshot({ path: 'screenshots\u002Fempty-projects.png', fullPage: true });\n```\n\nDocument: What does the system show when there is no data? Does it provide guidance? Is the UI broken?\n\n### 5.2 Long Content\n\n```javascript\n\u002F\u002F Submit forms with very long text\nawait page.fill('input[name=\"title\"]', 'A'.repeat(1000));\nawait page.screenshot({ path: 'screenshots\u002Flong-input.png' });\n```\n\nDocument: Does the system truncate? Overflow? Reject? Scroll?\n\n### 5.3 Responsive Layout\n\n```javascript\n\u002F\u002F Test at different viewport sizes\nconst viewports = [\n  { width: 375, height: 812, name: 'mobile' },     \u002F\u002F iPhone\n  { width: 768, height: 1024, name: 'tablet' },     \u002F\u002F iPad\n  { width: 1280, height: 800, name: 'laptop' },     \u002F\u002F Laptop\n  { width: 1920, height: 1080, name: 'desktop' },   \u002F\u002F Desktop\n];\n\nfor (const vp of viewports) {\n  await page.setViewportSize({ width: vp.width, height: vp.height });\n  await page.screenshot({ path: `screenshots\u002Fresponsive-${vp.name}.png`, fullPage: true });\n}\n```\n\nDocument: Does the layout adapt? Is navigation accessible at mobile sizes? Are interactive elements reachable?\n\n### 5.4 Accessibility\n\n```javascript\n\u002F\u002F Check for keyboard navigation\nawait page.keyboard.press('Tab');\nawait page.screenshot({ path: 'screenshots\u002Fkeyboard-focus-1.png' });\nawait page.keyboard.press('Tab');\nawait page.screenshot({ path: 'screenshots\u002Fkeyboard-focus-2.png' });\n\n\u002F\u002F Check for skip links\nawait page.keyboard.press('Tab'); \u002F\u002F First tab should focus skip-to-content if present\n```\n\nDocument: Is keyboard navigation possible? Are focus indicators visible? Is there a skip-to-content link?\n\n## Screenshot Naming Convention\n\nAll screenshots follow this naming scheme:\n\n```\n{phase}-{context}-{description}.png\n```\n\n| Phase Prefix | Usage |\n|-------------|-------|\n| `001-` through `099-` | Phase 1: Initial reconnaissance |\n| `nav-NNN-` | Phase 2: Navigation mapping |\n| `form-{page}-` | Phase 3: Interaction documentation |\n| `step-NN-` | Phase 4: Flow documentation (within flow subdirectory) |\n| `edge-{type}-` | Phase 5: Edge case exploration |\n\n### Directory Structure\n\n```\nworkspace\u002Fraw\u002Fruntime\u002Fvisual\u002F\n    screenshots\u002F                    # All screenshots (flat, except flows)\n        001-landing.png\n        nav-001-dashboard.png\n        nav-002-analytics.png\n        form-profile-before.png\n        form-profile-after.png\n        edge-empty-projects.png\n        edge-responsive-mobile.png\n    flows\u002F                          # End-to-end flow documentation\n        onboarding\u002F\n            step-01-landing.png\n            step-02-signup.png\n            flow.md\n        core-workflow\u002F\n            step-01-start.png\n            step-02-result.png\n            flow.md\n    navigation-map.md               # Complete navigation tree with screenshot refs\n    page-inventory.md               # Per-page interactive element inventory\n    visual-behaviors.md             # Behavioral claims synthesized from all observations\n```\n\n## Provenance Rules\n\n### Source Type\n\nAll claims from visual exploration use `source=visual-observation`:\n\n```markdown\n- The dashboard displays a \"No projects\" message with a \"Create Project\" call-to-action when the project list is empty\n  \u003C!-- cite: source=visual-observation, ref=screenshots\u002Fedge-empty-projects.png, confidence=confirmed, agent=visual-explorer -->\n```\n\n### Confidence Levels\n\n- **confirmed** -- the behavior is directly visible in a screenshot or demonstrated by a before\u002Fafter screenshot pair\n- **inferred** -- the behavior is implied by UI elements but not directly observed (e.g., a \"Delete\" button exists but was not clicked)\n- **assumed** -- the behavior is guessed from visual cues (e.g., a grayed-out button is assumed to be disabled)\n\n### Cite As You Go\n\nEvery behavioral claim gets an inline citation immediately after the claim. The `ref` field should be the screenshot path.\n\n## Rules\n\n1. **Screenshots are RAW** -- visual captures of proprietary UI go to `workspace\u002Fraw\u002F`. They never reach `workspace\u002Foutput\u002F` or `workspace\u002Fpublic\u002F`.\n2. **Screenshot everything** -- when in doubt, take a screenshot. Storage is cheap. Missing a state transition screenshot cannot be recovered without re-running the exploration.\n3. **Before AND after** -- every interaction must have a before screenshot and an after screenshot. The behavioral claim is the DIFFERENCE between the two.\n4. **Wait for stability** -- after every action, wait for the page to settle before screenshotting. Use `networkidle` or explicit waits for dynamic content.\n5. **Do not invent interactions** -- only interact with elements that are visible and actionable. If a button is disabled, document that it is disabled. Do not force-enable it.\n6. **Auth gates are data** -- if a page requires authentication you do not have, that gate is a behavioral claim worth documenting.\n7. **Cite as you go** -- every behavioral claim gets an inline `\u003C!-- cite: -->` comment immediately after the claim. Never defer citation to a later step.\n8. **Full page screenshots** -- use `fullPage: true` to capture below-the-fold content. Viewport-only screenshots miss content.\n","# Visual Exploration Methodology\n\nDocument behavioral intelligence through browser automation and visual observation. Every screenshot is an empirical observation. Every interaction sequence is a behavioral flow. This mode captures what the user SEES and what the system DOES in response to user actions.\n\n## When to Use This Mode\n\nVisual exploration activates when:\n- The target is a web application with a UI\n- The discovery inventory identifies a running web UI accessible via HTTP\n- A running instance of the target is accessible via HTTP\n\nThis mode requires a running instance of the target application and browser automation capabilities (Playwright, Puppeteer, or equivalent). All output is **RAW** (screenshots and flow documentation capture the target's UI in detail).\n\nTargets without a web UI get no useful signal from this mode — skip it.\n\n## Why Visual Exploration Matters\n\nSource code tells you what the system CAN do. Tests tell you what the system MUST do. Visual exploration tells you what the system LOOKS LIKE while doing it. This is behavioral intelligence that no other mode captures:\n\n- **Affordances** -- what actions does the UI present as available?\n- **Feedback** -- how does the system communicate success, failure, and state changes?\n- **Navigation structure** -- how are features organized from the user's perspective?\n- **State representation** -- how does the UI change as the system's state changes?\n- **Accessibility** -- is the interface usable with keyboard, screen reader, or at different sizes?\n\n## Phase 1: Initial Reconnaissance\n\n**Goal:** Navigate to the root URL, capture the landing page, and identify the application type, auth requirements, and available affordances.\n\n### 1.1 Landing Page Capture\n\n```bash\n# Navigate to root URL and screenshot\n# (Using Playwright as reference; adapt to available browser automation)\n```\n\n```javascript\nconst page = await browser.newPage();\nawait page.goto(ROOT_URL, { waitUntil: 'networkidle' });\nawait page.screenshot({ path: 'workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fscreenshots\u002F001-landing.png', fullPage: true });\n```\n\n### 1.2 Initial Assessment\n\nDocument:\n- **Application type** -- SPA, MPA, dashboard, wizard, documentation site, etc.\n- **Authentication** -- does the landing page require login? Is there a signup flow?\n- **Primary navigation** -- top nav, sidebar, hamburger menu, tabs, breadcrumbs?\n- **Visible affordances** -- buttons, links, forms, search bars, dropdowns visible on the landing page\n- **Branding and layout** -- header, footer, sidebar, content area dimensions\n\n### 1.3 Auth Handling\n\nIf authentication is required:\n1. Screenshot the login page\n2. Document the auth method (username\u002Fpassword, OAuth, SSO, API key)\n3. If test credentials are available, authenticate and screenshot the post-login state\n4. If no credentials are available, document the gate and explore only the unauthenticated surface\n\n## Phase 2: Navigation Mapping\n\n**Goal:** Click every navigation element, screenshot each page, and document the complete navigation tree.\n\n### 2.1 Identify Navigation Elements\n\n```javascript\n\u002F\u002F Extract all navigation links\nconst navLinks = await page.$$eval('nav a, [role=\"navigation\"] a, .sidebar a, .menu a, header a',\n  links => links.map(a => ({ text: a.textContent.trim(), href: a.href }))\n);\n```\n\n### 2.2 Systematic Traversal\n\nFor each navigation element:\n1. Click the element (or navigate to the href)\n2. Wait for the page to settle (network idle or DOM stable)\n3. Take a full-page screenshot\n4. Record the URL, page title, and visible heading\n5. Identify sub-navigation elements on the new page\n\n```javascript\nfor (const [index, link] of navLinks.entries()) {\n  await page.goto(link.href, { waitUntil: 'networkidle' });\n  await page.screenshot({\n    path: `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fscreenshots\u002Fnav-${String(index).padStart(3, '0')}-${slugify(link.text)}.png`,\n    fullPage: true\n  });\n}\n```\n\n### 2.3 Navigation Tree Output\n\nWrite to `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fnavigation-map.md`:\n\n```markdown\n## Navigation Tree\n\n- **Home** (`\u002F`) -- screenshot: 001-landing.png\n  - **Dashboard** (`\u002Fdashboard`) -- screenshot: nav-001-dashboard.png\n    - **Analytics** (`\u002Fdashboard\u002Fanalytics`) -- screenshot: nav-002-analytics.png\n    - **Reports** (`\u002Fdashboard\u002Freports`) -- screenshot: nav-003-reports.png\n  - **Settings** (`\u002Fsettings`) -- screenshot: nav-004-settings.png\n    - **Profile** (`\u002Fsettings\u002Fprofile`) -- screenshot: nav-005-profile.png\n    - **Billing** (`\u002Fsettings\u002Fbilling`) -- screenshot: nav-006-billing.png\n```\n\n## Phase 3: Interaction Documentation\n\n**Goal:** For each page, identify interactive elements, interact with them, and capture before\u002Fafter screenshots documenting the system's feedback.\n\n### 3.1 Element Inventory\n\nFor each page, identify:\n- **Forms** -- text inputs, selects, checkboxes, radio buttons, textareas, file uploads\n- **Buttons** -- submit buttons, action buttons, toggle buttons\n- **Dropdowns** -- select menus, custom dropdown components\n- **Modals\u002Fdialogs** -- elements that trigger overlay content\n- **Expandable sections** -- accordions, collapsible panels, \"show more\" links\n- **Interactive widgets** -- date pickers, sliders, color pickers, drag-and-drop zones\n\n```javascript\nconst interactiveElements = await page.$$eval(\n  'button, input, select, textarea, [role=\"button\"], [role=\"tab\"], [onclick], a[href=\"#\"]',\n  els => els.map(el => ({\n    tag: el.tagName,\n    type: el.type || '',\n    text: el.textContent?.trim().substring(0, 50),\n    id: el.id,\n    name: el.name,\n    ariaLabel: el.getAttribute('aria-label'),\n  }))\n);\n```\n\n### 3.2 Interaction Capture\n\nFor each interactive element:\n\n1. **Before screenshot** -- capture the page state before interaction\n2. **Interact** -- click, type, select, toggle\n3. **After screenshot** -- capture the page state after interaction\n4. **Document feedback** -- what changed? Error message? Success toast? New content?\n\n```javascript\n\u002F\u002F Example: form submission\nawait page.screenshot({ path: `screenshots\u002Fform-${name}-before.png` });\nawait page.fill('input[name=\"email\"]', 'test@example.com');\nawait page.click('button[type=\"submit\"]');\nawait page.waitForTimeout(1000); \u002F\u002F Allow time for feedback\nawait page.screenshot({ path: `screenshots\u002Fform-${name}-after.png` });\n```\n\n### 3.3 Page Inventory Output\n\nWrite to `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fpage-inventory.md`:\n\n```markdown\n## Page: Settings > Profile\n\n**URL:** `\u002Fsettings\u002Fprofile`\n**Screenshot:** nav-005-profile.png\n\n### Interactive Elements\n\n| Element | Type | Action Taken | Feedback | Screenshots |\n|---------|------|-------------|----------|-------------|\n| Display Name | text input | Typed \"Test User\" | No immediate feedback | form-profile-before.png, form-profile-after.png |\n| Save | submit button | Clicked | Success toast: \"Profile updated\" | form-profile-save.png |\n| Avatar | file upload | Uploaded test.png | Preview updated | form-profile-avatar.png |\n```\n\n## Phase 4: Flow Documentation\n\n**Goal:** Document end-to-end user flows with screenshots at every state transition.\n\n### 4.1 Core Flows to Document\n\nDocument these flows whenever they exist:\n\n| Flow | Description | Priority |\n|------|-------------|----------|\n| **Onboarding** | First-time user experience from signup to first use | High |\n| **Core workflow** | The primary task the application enables | High |\n| **Error recovery** | What happens when something goes wrong | High |\n| **Settings\u002Fconfiguration** | How users customize the application | Medium |\n| **Search\u002Ffilter** | How users find content within the application | Medium |\n| **CRUD operations** | Create, read, update, delete for primary entities | Medium |\n| **Authentication flow** | Login, logout, password reset, session expiry | Medium |\n| **Export\u002Fimport** | How data moves in and out of the application | Lower |\n\n### 4.2 Flow Documentation Format\n\nFor each flow, create a directory under `workspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fflows\u002F`:\n\n```\nworkspace\u002Fraw\u002Fruntime\u002Fvisual\u002Fflows\u002F\n    onboarding\u002F\n        step-01-landing.png\n        step-02-signup-form.png\n        step-03-email-verification.png\n        step-04-profile-setup.png\n        step-05-first-dashboard.png\n        flow.md\n    core-workflow\u002F\n        step-01-start.png\n        step-02-input.png\n        step-03-processing.png\n        step-04-result.png\n        flow.md\n```\n\nEach `flow.md` follows this format:\n\n```markdown\n# Flow: Onboarding\n\n## Steps\n\n### Step 1: Landing Page\n**Screenshot:** step-01-landing.png\n**URL:** `\u002F`\n**Action:** Click \"Sign Up\" button\n**State:** Unauthenticated, no account\n\n### Step 2: Signup Form\n**Screenshot:** step-02-signup-form.png\n**URL:** `\u002Fsignup`\n**Action:** Fill email, password, click \"Create Account\"\n**State:** Unauthenticated, form displayed\n**Feedback:** Form validation on blur for email format\n\n### Step 3: Email Verification\n**Screenshot:** step-03-email-verification.png\n**URL:** `\u002Fverify`\n**Action:** (requires email access -- documented but not tested)\n**State:** Account created, not verified\n**Feedback:** \"Check your email\" message displayed\n\n## Behavioral Claims\n\n- New accounts require email verification before first use\n  \u003C!-- cite: source=visual-observation, ref=flows\u002Fonboarding\u002Fstep-03-email-verification.png, confidence=confirmed, agent=visual-explorer -->\n```\n\n## Phase 5: Edge Cases\n\n**Goal:** Explore edge cases that reveal the system's behavior at its boundaries.\n\n### 5.1 Empty States\n\n```javascript\n\u002F\u002F Navigate to pages likely to have empty states (lists, dashboards, search results)\n\u002F\u002F Screenshot each empty state\nawait page.goto('\u002Fprojects'); \u002F\u002F Assuming no projects exist\nawait page.screenshot({ path: 'screenshots\u002Fempty-projects.png', fullPage: true });\n```\n\nDocument: What does the system show when there is no data? Does it provide guidance? Is the UI broken?\n\n### 5.2 Long Content\n\n```javascript\n\u002F\u002F Submit forms with very long text\nawait page.fill('input[name=\"title\"]', 'A'.repeat(1000));\nawait page.screenshot({ path: 'screenshots\u002Flong-input.png' });\n```\n\nDocument: Does the system truncate? Overflow? Reject? Scroll?\n\n### 5.3 Responsive Layout\n\n```javascript\n\u002F\u002F Test at different viewport sizes\nconst viewports = [\n  { width: 375, height: 812, name: 'mobile' },     \u002F\u002F iPhone\n  { width: 768, height: 1024, name: 'tablet' },     \u002F\u002F iPad\n  { width: 1280, height: 800, name: 'laptop' },     \u002F\u002F Laptop\n  { width: 1920, height: 1080, name: 'desktop' },   \u002F\u002F Desktop\n];\n\nfor (const vp of viewports) {\n  await page.setViewportSize({ width: vp.width, height: vp.height });\n  await page.screenshot({ path: `screenshots\u002Fresponsive-${vp.name}.png`, fullPage: true });\n}\n```\n\nDocument: Does the layout adapt? Is navigation accessible at mobile sizes? Are interactive elements reachable?\n\n### 5.4 Accessibility\n\n```javascript\n\u002F\u002F Check for keyboard navigation\nawait page.keyboard.press('Tab');\nawait page.screenshot({ path: 'screenshots\u002Fkeyboard-focus-1.png' });\nawait page.keyboard.press('Tab');\nawait page.screenshot({ path: 'screenshots\u002Fkeyboard-focus-2.png' });\n\n\u002F\u002F Check for skip links\nawait page.keyboard.press('Tab'); \u002F\u002F First tab should focus skip-to-content if present\n```\n\nDocument: Is keyboard navigation possible? Are focus indicators visible? Is there a skip-to-content link?\n\n## Screenshot Naming Convention\n\nAll screenshots follow this naming scheme:\n\n```\n{phase}-{context}-{description}.png\n```\n\n| Phase Prefix | Usage |\n|-------------|-------|\n| `001-` through `099-` | Phase 1: Initial reconnaissance |\n| `nav-NNN-` | Phase 2: Navigation mapping |\n| `form-{page}-` | Phase 3: Interaction documentation |\n| `step-NN-` | Phase 4: Flow documentation (within flow subdirectory) |\n| `edge-{type}-` | Phase 5: Edge case exploration |\n\n### Directory Structure\n\n```\nworkspace\u002Fraw\u002Fruntime\u002Fvisual\u002F\n    screenshots\u002F                    # All screenshots (flat, except flows)\n        001-landing.png\n        nav-001-dashboard.png\n        nav-002-analytics.png\n        form-profile-before.png\n        form-profile-after.png\n        edge-empty-projects.png\n        edge-responsive-mobile.png\n    flows\u002F                          # End-to-end flow documentation\n        onboarding\u002F\n            step-01-landing.png\n            step-02-signup.png\n            flow.md\n        core-workflow\u002F\n            step-01-start.png\n            step-02-result.png\n            flow.md\n    navigation-map.md               # Complete navigation tree with screenshot refs\n    page-inventory.md               # Per-page interactive element inventory\n    visual-behaviors.md             # Behavioral claims synthesized from all observations\n```\n\n## Provenance Rules\n\n### Source Type\n\nAll claims from visual exploration use `source=visual-observation`:\n\n```markdown\n- The dashboard displays a \"No projects\" message with a \"Create Project\" call-to-action when the project list is empty\n  \u003C!-- cite: source=visual-observation, ref=screenshots\u002Fedge-empty-projects.png, confidence=confirmed, agent=visual-explorer -->\n```\n\n### Confidence Levels\n\n- **confirmed** -- the behavior is directly visible in a screenshot or demonstrated by a before\u002Fafter screenshot pair\n- **inferred** -- the behavior is implied by UI elements but not directly observed (e.g., a \"Delete\" button exists but was not clicked)\n- **assumed** -- the behavior is guessed from visual cues (e.g., a grayed-out button is assumed to be disabled)\n\n### Cite As You Go\n\nEvery behavioral claim gets an inline citation immediately after the claim. The `ref` field should be the screenshot path.\n\n## Rules\n\n1. **Screenshots are RAW** -- visual captures of proprietary UI go to `workspace\u002Fraw\u002F`. They never reach `workspace\u002Foutput\u002F` or `workspace\u002Fpublic\u002F`.\n2. **Screenshot everything** -- when in doubt, take a screenshot. Storage is cheap. Missing a state transition screenshot cannot be recovered without re-running the exploration.\n3. **Before AND after** -- every interaction must have a before screenshot and an after screenshot. The behavioral claim is the DIFFERENCE between the two.\n4. **Wait for stability** -- after every action, wait for the page to settle before screenshotting. Use `networkidle` or explicit waits for dynamic content.\n5. **Do not invent interactions** -- only interact with elements that are visible and actionable. If a button is disabled, document that it is disabled. Do not force-enable it.\n6. **Auth gates are data** -- if a page requires authentication you do not have, that gate is a behavioral claim worth documenting.\n7. **Cite as you go** -- every behavioral claim gets an inline `\u003C!-- cite: -->` comment immediately after the claim. Never defer citation to a later step.\n8. **Full page screenshots** -- use `fullPage: true` to capture below-the-fold content. Viewport-only screenshots miss content.\n",{"name":5,"description":7},"main","728894f3f85fbd4e788f015ddfd72bb6155d5dd5","plugins\u002Fmoe-backstory\u002F.opencode\u002Fskills\u002Fvisual-exploration",1,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe","https:\u002F\u002Fgithub.com\u002Fzak-keown","2026-09-05 06:29:06",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add zak-keown\u002Fmoe --skill visual-exploration",{"tool":36,"command":37},"bunx","bunx skills add zak-keown\u002Fmoe --skill visual-exploration",{"tool":39,"command":40},"pnpm","pnpm dlx skills add zak-keown\u002Fmoe --skill visual-exploration","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe-backstory\u002F.opencode\u002Fskills\u002Fvisual-exploration\u002FSKILL.md",true,[],[45,52,59,66,73,80],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":18},"01M7C1Z5362Y8NVRZ69HWP0ZD0","write-skill","Use when creating new skills, editing existing skills, or verifying skills work before deployment","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-skill\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":18},"01M7C1YZ9966WK9R6ZRPCGWF3G","write-plan","Use when you have a spec or requirements for a multi-step task, before touching code","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-plan\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":18},"01M7C1YMR4SJD7BPBBPXZCVGVH","write-clearly","Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Makes your writing clearer, stronger, and more professional.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-clearly\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":18},"01M7C1YH9805YSM0N20VQAVSXF","windows-vm","Create, manage, or connect to a headless Windows 11 VM running in Docker with SSH access. Use when the user wants to spin up, stop, restart, or SSH into a Windows VM.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwindows-vm\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M7C1YD87SCNEJ2C9B8HFR2NQ","verify-completion","Use when about to claim work is complete, fixed, or passing, before committing or creating MRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fverify-completion\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"01M7C1YARDEHT0TJ9BRG30ZCF4","using-moe","Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fusing-moe\u002FSKILL.md",[],[]]