Improve MASTER-PROMPT.md examples section clarity

- Create dedicated 'How Auto-Triggering Works: Practical Examples' section
- Add clear NOTE explaining these are illustrative examples
- Consolidate 5 examples showing WITHOUT vs WITH scenarios
- Add visual separation with headers and dividers
- Include summary table comparing without vs with auto-triggering
- Remove duplicate examples scattered throughout document
- Make it immediately clear these are workflow examples

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
uroma
2026-01-15 16:37:11 +00:00
Unverified
parent 34499743fe
commit f82524b86d

View File

@@ -171,9 +171,15 @@ cp -r /tmp/contains-studio-agents/agents/* ~/.claude/agents/
- Comprehensive styling guidelines - Comprehensive styling guidelines
- Example: "Create a pricing page" → ui-ux-pro-max applies professional design patterns - Example: "Create a pricing page" → ui-ux-pro-max applies professional design patterns
### Real-Life Example: Before vs After ---
**BEFORE (Without agents):** ## 📖 How Auto-Triggering Works: Practical Examples
> **NOTE:** The examples below illustrate the difference between working without auto-triggering agents vs with auto-triggering agents. These are simplified scenarios showing how the PROACTIVELY system improves your workflow.
### Example 1: Adding OAuth Authentication
**WITHOUT Auto-Triggering Agents:**
``` ```
You: Help me add OAuth to my app You: Help me add OAuth to my app
Claude: [Writes code, but no tests] Claude: [Writes code, but no tests]
@@ -182,7 +188,7 @@ Claude: [Fixes bugs]
You: [Deployment issues] You: [Deployment issues]
``` ```
**AFTER (With auto-triggering agents):** **WITH Auto-Triggering Agents:**
``` ```
You: Help me add OAuth to my app You: Help me add OAuth to my app
Claude: [Writes code] Claude: [Writes code]
@@ -201,6 +207,131 @@ Claude (as whimsy-injector): Adding delightful touches to OAuth UI...
Ready to deploy! Ready to deploy!
``` ```
**Key Difference:** Auto-triggering agents proactively handle tasks (testing, UX refinement) that you'd otherwise need to remember and do manually.
---
### Example 2: UI Component Creation (ui-ux-pro-max)
**WITHOUT Auto-Triggering:**
```
You: Create a modal for user settings
Claude: [Creates basic modal without accessibility]
[Missing: ARIA labels, focus trapping, keyboard support]
[Result: Modal doesn't work with screen readers]
```
**WITH Auto-Triggering (ui-ux-pro-max):**
```
You: Create a modal for user settings
[ui-ux-pro-max PROACTIVELY triggers]
Claude (as ui-ux-pro-max): I'll create a professional, accessible modal...
✓ Proper ARIA labels and roles
✓ Focus trap implementation
✓ Escape key and backdrop click handling
✓ Smooth animations (150-300ms)
✓ Mobile-responsive sizing
✓ WCAG 2.1 AA compliant colors
[Result: Modal works perfectly for all users]
```
---
### Example 3: Error Diagnosis from Screenshots
**WITHOUT Vision Tools:**
```
You: [Paste error screenshot]
Claude: I can't see images. Please paste the text.
You: [Types error manually]
Claude: Let me help...
[Misses visual context like line numbers, file paths shown in screenshot]
```
**WITH Vision Tools (diagnose_error):**
```
You: [Upload error-screenshot.png]
You: Analyze this error
Claude: [Using diagnose_error tool]
I can see this is a TypeError in user.service.ts at line 42.
The error "Cannot read property 'id' of undefined" occurs because
the user object is null when the API call fails.
The stack trace shows it originated in the login function.
Fix: Add null check after API call.
[Provides precise fix with visual context]
```
---
### Example 4: Understanding Large Codebases
**WITHOUT Token-Efficient Analysis:**
```
You: Explain how authentication works in this 100K-line codebase
Claude: [Reading files... token limit exceeded...]
Cannot fit entire codebase in context.
```
**WITH TLDR (95% Token Reduction):**
```
You: Explain how authentication works
Claude: [Using tldr context]
Analysis complete in 2 seconds, 12K tokens (vs 104K raw)
Authentication Flow:
1. login() (app/auth.py:45) - Entry point
→ validate_credentials() (app/auth.py:78)
→ hash_password() (app/crypto.py:23)
→ create_session() (app/session.py:12)
→ redirect_to_dashboard() (app/routes.py:156)
Key dependencies:
- Uses bcrypt for hashing
- Session stored in Redis
- 30-minute timeout
[Complete understanding with 89% fewer tokens]
```
---
### Example 5: Web Research for Current Information
**WITHOUT Web Search:**
```
You: What's the latest React version?
Claude: [Training cutoff knowledge] React 18 was released in 2022.
You: [Outdated information]
```
**WITH Web Search (web-search-prime):**
```
You: What's the latest React version and features?
Claude: [Using web-search-prime]
Searching current web...
React 19 was released in December 2024 with new features:
- Server Actions
- Enhanced useTransition
- New Suspense features
[Fully up-to-date information]
```
---
### Summary: The Auto-Triggering Advantage
| Aspect | Without Auto-Triggering | With Auto-Triggering |
|--------|------------------------|---------------------|
| **Testing** | Manual, forget to do it | Automatic after code changes |
| **UX Polish** | Basic, inconsistent | Professional, accessible |
| **Error Analysis** | Type text manually | Upload screenshot, instant diagnosis |
| **Large Codebases** | Token limits, incomplete | 95% reduction, complete understanding |
| **Research** | Outdated knowledge | Real-time web search |
| **Your Role** | Remember everything | Focus on core logic, agents handle rest |
**Bottom Line:** Auto-triggering agents handle the "should-do" tasks (testing, UX polish, documentation) that you know you should do but often forget or skip due to time constraints.
---
--- ---
## Step 1.5: Install UI/UX Pro Max Agent (PROACTIVELY Auto-Triggers) ## Step 1.5: Install UI/UX Pro Max Agent (PROACTIVELY Auto-Triggers)
@@ -235,55 +366,11 @@ cp /tmp/ui-ux-pro-max-skill/.claude/skills/ui-ux-pro-max/SKILL.md ~/.claude/skil
- Accessibility standards (contrast ratios, keyboard nav, ARIA labels) - Accessibility standards (contrast ratios, keyboard nav, ARIA labels)
- Professional design patterns (no emoji icons, proper hover states) - Professional design patterns (no emoji icons, proper hover states)
### Real-Life Examples: Before vs After **Key Features:**
- Automatic UI/UX review during design work
**Example: UI Component Creation** - Professional design patterns applied proactively
- Accessibility-first approach (WCAG 2.1 AA/AAA)
**BEFORE:** - Comprehensive styling guidelines
```
You: Create a modal for user settings
Claude: [Creates basic modal without accessibility]
[Missing: ARIA labels, focus trapping, keyboard support]
[Result: Modal doesn't work with screen readers]
```
**AFTER:**
```
You: Create a modal for user settings
[ui-ux-pro-max PROACTIVELY triggers]
Claude (as ui-ux-pro-max): I'll create a professional, accessible modal...
✓ Proper ARIA labels and roles
✓ Focus trap implementation
✓ Escape key and backdrop click handling
✓ Smooth animations (150-300ms)
✓ Mobile-responsive sizing
✓ WCAG 2.1 AA compliant colors
[Result: Modal works perfectly for all users]
```
**Example: UI Review**
**BEFORE:**
```
You: The dashboard looks messy
Claude: Let me add some spacing...
[Generic fixes, doesn't address root issues]
```
**AFTER:**
```
You: The dashboard looks messy
[ui-ux-pro-max PROACTIVELY triggers]
Claude (as ui-ux-pro-max): Analyzing dashboard for professional issues...
Issues found:
✗ Inconsistent spacing (mixing 4px and 8px grids)
✗ Low contrast text (slate-400 unreadable in light mode)
✗ Missing cursor pointers on cards
✗ No hover feedback on interactive elements
✗ z-index conflicts between components
[Provides specific fixes for each issue]
```
--- ---
@@ -308,58 +395,11 @@ npx @z_ai/mcp-server --help
7. **analyze_data_viz** - Extract insights from dashboards/charts 7. **analyze_data_viz** - Extract insights from dashboards/charts
8. **understand_diagram** - Understand technical diagrams (UML, flowcharts) 8. **understand_diagram** - Understand technical diagrams (UML, flowcharts)
### Real-Life Examples: Before vs After **Key Features:**
- Screenshot-to-code conversion
**Example 1: Error Diagnosis** - Error diagnosis from images
- Video and diagram understanding
**BEFORE:** - UI comparison and analysis
```
You: [Paste error screenshot]
Claude: I can't see images. Please paste the text.
You: [Types error manually]
Claude: Let me help...
[Misses visual context like line numbers, file paths shown in screenshot]
```
**AFTER:**
```
You: [Upload error-screenshot.png]
You: Analyze this error
Claude: [Using diagnose_error tool]
I can see this is a TypeError in user.service.ts at line 42.
The error "Cannot read property 'id' of undefined" occurs because
the user object is null when the API call fails.
The stack trace shows it originated in the login function.
Fix: Add null check after API call.
[Provides precise fix with visual context]
```
**Example 2: UI to Code**
**BEFORE:**
```
You: [Describes UI for 10 minutes]
Claude: [Creates approximation]
You: [Several iterations to match design]
```
**AFTER:**
```
You: [Upload design-mockup.png]
You: Implement this UI
Claude: [Using ui_to_artifact tool]
Analyzing design...
- Header: Navigation bar with logo and links
- Hero: Large headline with gradient background
- Cards: 3-column grid with icons
- CTA: Gradient button with hover effect
[Creates exact match with Tailwind CSS]
✓ All spacing matches
✓ Colors extracted from design
✓ Icons identified and included
```
--- ---
@@ -379,49 +419,10 @@ npx @z_ai/coding-helper --help
2. **web-reader** - Convert any web page to clean markdown 2. **web-reader** - Convert any web page to clean markdown
3. **zread** - Read and analyze entire GitHub repositories 3. **zread** - Read and analyze entire GitHub repositories
### Real-Life Examples: Before vs After **Key Features:**
- Real-time web search for current information
**Example 1: Web Research** - Repository analysis without manual file pasting
- Web page to clean markdown conversion
**BEFORE:**
```
You: What's the latest React version?
Claude: [Training cutoff knowledge] React 18 was released in 2022.
You: [Outdated information]
```
**AFTER:**
```
You: What's the latest React version and features?
Claude: [Using web-search-prime]
Searching current web...
React 19 was released in December 2024 with new features:
- Server Actions
- Enhanced useTransition
- New Suspense features
[Fully up-to-date information]
```
**Example 2: Repository Analysis**
**BEFORE:**
```
You: Help me understand the Next.js codebase
Claude: [Cannot access external repositories]
You: [Manually paste files one by one]
```
**AFTER:**
```
You: Analyze the Next.js repository structure
Claude: [Using zread tool]
Reading vercel/next.js...
Repository has:
- 4500+ files
- Main packages: next, react, react-dom
- Key directories: packages/next, examples, test
[Provides comprehensive overview without manual file pasting]
```
--- ---
@@ -453,89 +454,11 @@ tldr warm .
- `dfg` - Data flow graph - `dfg` - Data flow graph
- And 12 more... - And 12 more...
### Real-Life Examples: Before vs After **Key Features:**
- 95% token reduction for large codebases
**Example 1: Understanding Large Codebase** - Semantic search by behavior (not just text)
- Program slicing for precise debugging
**BEFORE:** - Impact analysis for safe refactoring
```
You: Explain how authentication works in this 100K-line codebase
Claude: [Reading files... token limit exceeded...]
Cannot fit entire codebase in context.
```
**AFTER:**
```
You: Explain how authentication works
Claude: [Using tldr context]
Analysis complete in 2 seconds, 12K tokens (vs 104K raw)
Authentication Flow:
1. login() (app/auth.py:45) - Entry point
→ validate_credentials() (app/auth.py:78)
→ hash_password() (app/crypto.py:23)
→ create_session() (app/session.py:12)
→ redirect_to_dashboard() (app/routes.py:156)
Key dependencies:
- Uses bcrypt for hashing
- Session stored in Redis
- 30-minute timeout
[Complete understanding with 89% fewer tokens]
```
**Example 2: Semantic Search**
**BEFORE:**
```
You: Where's the JWT validation code?
Claude: [Searches for "JWT" or "validation"]
Found nothing (it's called "verify_access_token")
```
**AFTER:**
```
You: Find the JWT validation code
Claude: [Using tldr semantic search]
Searching by behavior "validate JWT tokens"...
Found: verify_access_token() in app/auth/middleware.py
This function:
- Validates JWT signature
- Checks expiration
- Extracts user claims
[Finds by behavior, not exact name match]
```
**Example 3: Debugging with Program Slicing**
**BEFORE:**
```
You: Why is user_id null on line 142?
Claude: [Reads entire 150-line function]
Let me analyze... it's hard to see the data flow...
[Ambiguous answer]
```
**AFTER:**
```
You: Why is user_id null on line 142?
Claude: [Using tldr slice]
Showing only lines affecting line 142...
3: user = db.get_user(username)
7: if user is None:
12: return None # ← BUG: Early return
28: token = create_token(user) # ← Skipped due to line 12
42: session.user_id = user.id # ← user is None here!
The issue: When user not found (line 7), function returns
None at line 12, skipping token creation. Line 42 assumes
user exists.
[Precise answer with only 6 relevant lines shown]
```
--- ---