Advanced Automation

Multi-source Information Collection → Intelligent Deduplication → Quality Scoring → Automatic Classification → Daily Report Generation → Lark Push Notification

# Install Information Collection Skills
clawhub install brave-search      # Web Search
clawhub install rss-reader        # RSS Subscription
clawhub install github-trending   # GitHub Trending

# Install Content Processing Skills
clawhub install content-analyzer  # Content Analysis
clawhub install text-summarizer   # Text Summarizer
clawhub install duplicate-checker # Deduplication Checker

{
  "sources": {
    "rss": [
      {
        "name": "阮一峰的网络日志",
        "url": "https://www.ruanyifeng.com/blog/atom.xml",
        "category": "Technology"
      },
      {
        "name": "少数派",
        "url": "https://sspai.com/feed",
        "category": "Productivity"
      }
    ],
    "github": {
      "trending": {
        "language": "python",
        "since": "daily"
      },
      "repos": [
        "openclaw/openclaw",
        "microsoft/vscode"
      ]
    },
    "keywords": [
      "OpenClaw",
      "AI tools",
      "Efficiency improvement",
      "Automation"
    ]
  },
  "filter": {
    "keywords": ["AI", "Automation", "Efficiency", "Tools"],
    "exclude": ["Advertisement", "Marketing", "Promotion"],
    "quality_threshold": 75,
    "max_items": 20
  },
  "schedule": {
    "time": "08:00",
    "timezone": "Asia/Shanghai"
  }
}

#!/bin/bash

# Daily Information Collection Script
DATE=$(date +%Y-%m-%d)
OUTPUT_DIR="$HOME/.openclaw/digests"
mkdir -p "$OUTPUT_DIR"

# 1. Collecting RSS information
echo "📰 Collecting RSS information..."
openclaw agent --message "Please use the rss-reader skill to collect RSS feeds configured in ~/.openclaw/info-sources.json and save them to $OUTPUT_DIR/rss-$DATE.json"

# 2. Collecting GitHub trending
echo "🔥 Collecting GitHub trending..."
openclaw agent --message "Please collect today's trending Python projects on GitHub and save them to $OUTPUT_DIR/github-$DATE.json"

# 3. Searching keywords
echo "🔍 Searching keywords..."
openclaw agent --message "Please search for information related to 'OpenClaw AI tools', with a maximum of 10 results, and save them to $OUTPUT_DIR/search-$DATE.json"

# 4. Merging and deduplicating
echo "🔄 Merging and deduplicating..."
openclaw agent --message "Please merge all information from $OUTPUT_DIR/*-$DATE.json, deduplicate it, and save it to $OUTPUT_DIR/merged-$DATE.json"

# 5. Content analysis and scoring
echo "📊 Content analysis..."
openclaw agent --message "Please analyze the content in $OUTPUT_DIR/merged-$DATE.json, score it, and save it to $OUTPUT_DIR/analyzed-$DATE.json"

# 6. Generating daily report
echo "📝 Generating daily report..."
openclaw agent --message "Please generate today's daily report based on the following information, categorized by technology, product, and industry, with each item including title, summary, link, and recommendation reason:
$(cat $OUTPUT_DIR/analyzed-$DATE.json)" \
  --output "$OUTPUT_DIR/digest-$DATE.md"

# 7. Pushing to Lark
echo "📤 Pushing to Lark..."
openclaw channels send feishu \
  --message "$(cat $OUTPUT_DIR/digest-$DATE.md)" \
  --title "📰 Daily News $DATE"

echo "✅ Daily report generation complete!"

# Add to crontab
crontab -e

# Execute daily at 8 AM
0 8 * * * /bin/bash ~/.openclaw/scripts/daily-digest.sh

{
  "sources": {
    "rss": [
      "https://news.ycombinator.com/rss",
      "https://www.reddit.com/r/programming/.rss",
      "https://dev.to/feed"
    ],
    "github": {
      "trending": ["python", "javascript", "go"],
      "topics": ["ai", "automation", "productivity"]
    },
    "twitter": {
      "accounts": ["@OpenAI", "@github", "@vercel"],
      "hashtags": ["#AI", "#DevTools"]
    }
  },
  "filter": {
    "keywords": ["AI", "Open Source", "Tools", "Frameworks"],
    "min_stars": 100,
    "quality_threshold": 80
  }
}

Task Input → Intelligent Analysis → Priority Sorting → Time Scheduling → Progress Tracking → Completion Review

clawhub install task-manager
clawhub install calendar-sync
clawhub install priority-analyzer
clawhub install time-estimator

{
  "priority_rules": {
    "urgent_keywords": ["Urgent", "Immediate", "Must today"],
    "important_keywords": ["Important", "Critical", "Core"],
    "deadline_weight": 0.4,
    "impact_weight": 0.3,
    "effort_weight": 0.3
  },
  "time_rules": {
    "work_hours": "09:00-18:00",
    "focus_time": "09:00-11:00",
    "meeting_time": "14:00-16:00",
    "break_interval": 90
  },
  "automation": {
    "auto_schedule": true,
    "auto_reminder": true,
    "auto_followup": true
  }
}

You: Help me manage today's tasks
OpenClaw: Okay, please tell me your task list for today.

You:
1. Complete project documentation (estimated 2 hours)
2. Code review (estimated 1 hour)
3. Team meeting (14:00-15:00)
4. Fix bug (urgent, estimated 3 hours)
5. Learn new technology (estimated 1 hour)

OpenClaw: Your tasks have been analyzed. Suggested execution order:

🔴 Priority 1 (Urgent & Important)
- Fix bug (09:00-12:00, 3 hours)
  Reason: Marked as urgent, scheduled during morning focus time

🟡 Priority 2 (Important but Not Urgent)
- Complete project documentation (13:00-14:00, 1 hour)
  Reason: Important task, to be completed before the meeting
- Code review (15:00-16:00, 1 hour)
  Reason: Requires focus, to be done after the meeting

🟢 Priority 3 (Not Urgent)
- Team meeting (14:00-15:00, fixed)
- Learn new technology (16:00-17:00, 1 hour)
  Reason: Flexible, scheduled for the afternoon

Synchronized with your calendar and reminders set. Do you need any adjustments?

# Create project
openclaw agent --message "Create project: OpenClaw Tutorial Optimization
Task list:
1. Fix Docker documentation (2 hours, completed)
2. Supplement Chapter 13 content (8 hours, in progress)
3. Delete duplicate content (4 hours, pending)
4. Unify code format (3 hours, pending)
5. Update cross-references (2 hours, pending)

Please help me create an execution plan"

# OpenClaw automatically generates Gantt chart and execution plan

Inspiration Collection → Material Organization → Outline Generation → Content Creation → Layout Optimization → Multi-platform Publishing

# Record inspiration anytime in Lark
You: Inspiration: OpenClaw can be used for automated testing
OpenClaw: Inspiration recorded under 'Content Creation' category
- Title: OpenClaw Automated Testing Application
- Keywords: Automation, Testing, OpenClaw
- Related materials: 3 related articles searched
- Suggested outline: Preliminary outline generated
- Estimated word count: 2000-3000 words

Do you want to start creating now?

#!/bin/bash

TOPIC=$1
OUTPUT_DIR="$HOME/.openclaw/content"
mkdir -p "$OUTPUT_DIR"

# 1. Searching for related materials
echo "🔍 Searching for related materials..."
openclaw skills run brave-search \
  --query "$TOPIC" \
  --max-results 10 \
  --output "$OUTPUT_DIR/research.json"

# 2. Generating outline
echo "📋 Generating outline..."
openclaw agent --message "Generate an article outline based on the following materials:
Topic: $TOPIC
Materials: $(cat $OUTPUT_DIR/research.json)
Requirements:
- Clear structure, 3-5 sections
- Include practical cases
- Word count 2000-3000 words" \
  --output "$OUTPUT_DIR/outline.md"

# 3. Creating content
echo "✍️ Creating content..."
openclaw agent --message "Create a complete article based on the outline:
$(cat $OUTPUT_DIR/outline.md)
Requirements:
- Easy-to-understand language
- Include code examples
- Suggest accompanying images" \
  --output "$OUTPUT_DIR/draft.md"

# 4. Optimizing layout
echo "🎨 Optimizing layout..."
openclaw agent --message "Optimize article layout:
$(cat $OUTPUT_DIR/draft.md)
Requirements:
- Add emoji
- Optimize heading levels
- Add quotes and callout boxes" \
  --output "$OUTPUT_DIR/final.md"

echo "✅ Content creation complete!"
echo "📄 File location: $OUTPUT_DIR/final.md"

# Create an article
bash ~/.openclaw/scripts/content-creation.sh "OpenClaw Automated Testing Practice"

# After 10 minutes, the article creation is complete.

# Monday: Generate this week's topics
openclaw agent --message "Analyze recent tech hot topics and generate 3 blog topics"

# Tuesday: Create the first article
bash ~/.openclaw/scripts/content-creation.sh "Topic 1"

# Wednesday: Review and optimize
openclaw agent --message "Review the article and suggest optimizations:
$(cat ~/.openclaw/content/final.md)"

# Thursday: Generate accompanying images
openclaw agent --message "Please generate accompanying images for the technical blog, style: minimalist tech"

# Friday: Multi-platform publishing
openclaw agent --message "Publish the article to:
- Personal blog
- Juejin
- CSDN
- Zhihu"

{
  "tracking": {
    "time_tracking": {
      "enabled": true,
      "categories": ["Work", "Study", "Rest", "Entertainment"],
      "auto_detect": true
    },
    "task_tracking": {
      "enabled": true,
      "sync_calendar": true,
      "track_completion": true
    },
    "focus_tracking": {
      "enabled": true,
      "pomodoro": 25,
      "break": 5
    }
  },
  "reporting": {
    "daily_summary": "20:00",
    "weekly_review": "Sunday 18:00",
    "monthly_report": "Last day 18:00"
  }
}

# Daily Summary
openclaw agent --message "Generate today's efficiency report"

# OpenClaw automatically generates:
📊 Today's Efficiency Report 2026-02-11

⏰ Time Allocation
- Work: 6.5 hours (Goal 8 hours)
- Study: 2 hours
- Rest: 1.5 hours

✅ Task Completion
- Planned tasks: 8
- Completed tasks: 7
- Completion rate: 87.5%

🎯 Focus Duration
- Deep work: 4 hours
- Pomodoros: 8
- Focus level: 85%

🤖 Automation Effect
- Automated tasks: 12
- Time saved: 2.5 hours
- Automation rate: 60%

💡 Improvement Suggestions
1. Increase deep work time to 5 hours
2. Reduce meeting time
3. Increase automation coverage to 70%

Skill A → Skill B → Skill C

        → Skill A →
Input →  → Skill B →  → Output
        → Skill C →

Input → Condition → Skill A (Condition 1)
                  → Skill B (Condition 2)

find-skills (Discover new Skills)
  ↓
ProactiveAgent (Predict demand)
  ↓
brave-search (Search for relevant information)
  ↓
content-analyzer (Analyze content)
  ↓
markdown-generator (Generate report)

web-clipper (Web Clipper)
  ↓
content-extractor (Extract main content)
  ↓
text-summarizer (Generate summary)
  ↓
tag-generator (Automatic tagging)
  ↓
notion-sync (Sync to Notion)

github-search (Search projects)
  ↓
code-analyzer (Analyze code)
  ↓
dependency-checker (Check dependencies)
  ↓
doc-generator (Generate documentation)
  ↓
knowledge-graph (Build knowledge graph)

Discover Learning Resources → Content Extraction → Knowledge Organization → Practice Exercises → Review and Summary

# Content Discovery
clawhub install github-trending
clawhub install course-finder
clawhub install paper-search

# Content Processing
clawhub install pdf-reader
clawhub install video-transcriber
clawhub install note-taker

# Knowledge Management
clawhub install flashcard-generator
clawhub install mind-map-creator
clawhub install spaced-repetition

{
  "workflow": "Automated Learning System",
  "steps": [
    {
      "name": "Discover Learning Resources",
      "skills": ["github-trending", "course-finder"],
      "config": {
        "topics": ["AI", "Automation", "Productivity Tools"],
        "quality_threshold": 80
      }
    },
    {
      "name": "Content Extraction",
      "skills": ["pdf-reader", "video-transcriber"],
      "config": {
        "extract_key_points": true,
        "generate_summary": true
      }
    },
    {
      "name": "Knowledge Organization",
      "skills": ["note-taker", "mind-map-creator"],
      "config": {
        "format": "markdown",
        "auto_categorize": true
      }
    },
    {
      "name": "Memory Reinforcement",
      "skills": ["flashcard-generator", "spaced-repetition"],
      "config": {
        "review_schedule": "1,3,7,15,30"
      }
    }
  ]
}

# Start learning workflow
openclaw workflow run learning

# Or start via dialogue
openclaw agent --message "I want to learn Python asynchronous programming, please help me create a study plan."

# OpenClaw automatically executes:
1. Search for Python asynchronous programming resources
   - GitHub trending projects: 5
   - High-quality tutorials: 3
   - Technical articles: 10

2. Extract core knowledge points
   - async/await syntax
   - Coroutine principles
   - Event loop
   - Common libraries: asyncio, aiohttp

3. Generate study notes
   - Concept explanations
   - Code examples
   - Practical projects

4. Create review flashcards
   - Generate 20 flashcards
   - Set review schedule

5. Recommend practical projects
   - Asynchronous web crawler
   - Asynchronous API service
   - Asynchronous task queue

Inspiration Collection → Material Search → Outline Generation → Content Creation → Image Generation → Layout Optimization → Multi-platform Publishing

# Inspiration Collection
idea-collector + trend-analyzer
  ↓
# Material Search
brave-search + content-scraper + image-search
  ↓
# Outline Generation
outline-generator + structure-optimizer
  ↓
# Content Creation
ai-writer + code-generator + example-creator
  ↓
# Image Generation
image-generator + image-optimizer
  ↓
# Layout Optimization
markdown-formatter + style-checker
  ↓
# Multi-platform Publishing
blog-publisher + social-media-poster

# Create content creation workflow
openclaw agent --message "I want to write an article about OpenClaw automation."

# OpenClaw automatically executes:

📝 Step 1: Analyze Hot Trends
- Search for content related to "OpenClaw automation"
- Analyze popular topics and user pain points
- Generate 3 topic directions

📚 Step 2: Collect Materials
- Search for related articles: 15
- Extract high-quality cases: 8
- Collect image materials: 20

📋 Step 3: Generate Outline
I. OpenClaw Automation Overview
II. 5 Practical Scenarios
III. Detailed Configuration Methods
IV. Performance Data Display
V. Pitfall Avoidance Guide

✍️ Step 4: Create Content
- Automatically generate content for each section
- Insert code examples
- Add practical cases

🎨 Step 5: Generate Images
- Cover image: Tech style
- Flowcharts: 3
- Effect comparison charts: 2

🔧 Step 6: Optimize Layout
- Add emoji
- Optimize heading levels
- Add quote boxes

📤 Step 7: Multi-platform Publishing
- Personal blog: Published
- Juejin: Published
- Zhihu: Published
- Official Account: Draft generated

✅ Creation complete! Total time: 15 minutes

Data Collection → Data Cleaning → Data Analysis → Visualization → Report Generation

# Data Collection
api-connector + web-scraper + database-query
  ↓
# Data Cleaning
data-cleaner + duplicate-remover + format-converter
  ↓
# Data Analysis
statistical-analyzer + trend-detector + anomaly-finder
  ↓
# Visualization
chart-generator + dashboard-creator
  ↓
# Report Generation
report-writer + insight-summarizer

# Analyze GitHub project data
openclaw agent --message "Analyze the growth trend of the OpenClaw project"

# OpenClaw automatically executes:

📊 Step 1: Collect Data
- Star history: 1000 records
- Fork history: 500 records
- Issue statistics: 200
- PR statistics: 150

🧹 Step 2: Data Cleaning
- Remove duplicate data
- Fill missing values
- Standardize format

📈 Step 3: Data Analysis
- Growth trend: Average monthly growth of 15%
- Activity: Peak during weekdays
- User profile: 70% developers
- Popular features: Skills are most popular

📉 Step 4: Generate Charts
- Star growth curve
- User activity heatmap
- Feature usage pie chart
- Issue classification bar chart

📝 Step 5: Generate Report
《OpenClaw Project Data Analysis Report》
- Core metrics
- Growth trends
- User insights
- Improvement suggestions

✅ Analysis complete!

# Bad practice: One Skill does everything
clawhub install all-in-one-tool

# Good practice: Combination of multiple specialized Skills
clawhub install data-collector
clawhub install data-analyzer
clawhub install report-generator

{
  "workflow": "Content Creation",
  "error_handling": {
    "retry": 3,
    "fallback": "use_default",
    "notification": true
  },
  "steps": [
    {
      "name": "Search materials",
      "skill": "brave-search",
      "on_error": "use_cached_data"
    },
    {
      "name": "Generate content",
      "skill": "ai-writer",
      "on_error": "use_template"
    }
  ]
}

# Sequential execution (slow)
Search materials → Generate outline → Create content → Generate images
Total time: 20 minutes

# Parallel execution (fast)
Search materials ─┬→ Generate outline → Create content
          └→ Generate images
Total time: 12 minutes

{
  "data_format": {
    "input": {
      "type": "json",
      "schema": "standard_v1"
    },
    "output": {
      "type": "json",
      "schema": "standard_v1"
    }
  }
}

# Use a locking mechanism
openclaw config set skills.lock true

# Or execute sequentially
openclaw workflow run --mode sequential

# Limit concurrency
openclaw config set skills.max_concurrent 3

# Set priority
openclaw config set skills.priority '{
  "critical": ["task-manager"],
  "high": ["content-creator"],
  "normal": ["data-analyzer"]
}'

{
  "dependencies": {
    "skill-b": {
      "requires": ["skill-a"],
      "on_missing": "skip"
    }
  }
}

Fragmented Knowledge → Structured → Relational → Visualized → Searchable → Reusable

OpenClaw (Core Concept)
  ├─ Includes → Skills (Sub-concept)
  │   ├─ Includes → find-skills
  │   ├─ Includes → ProactiveAgent
  │   └─ Includes → brave-search
  ├─ Applied to → Automation (Application Scenario)
  │   ├─ Includes → Information Collection
  │   ├─ Includes → Task Management
  │   └─ Includes → Content Creation
  └─ Related to → AI Tools (Related Concept)
      ├─ Includes → ChatGPT
      ├─ Includes → Claude
      └─ Includes → Copilot

# Extract knowledge points from notes
clawhub install note-parser
openclaw agent --message "Please use the note-parser skill to extract knowledge points from ~/.openclaw/notes and save them to ~/.openclaw/knowledge/entities.json"

# Extract from browsing history
clawhub install browser-history-analyzer
openclaw agent --message "Please use the browser-history-analyzer skill to analyze the last 30 days of browsing history and save it to ~/.openclaw/knowledge/topics.json"

# Extract from chat history
openclaw agent --message "Analyze my recent conversations and extract key knowledge points"

# Annotate knowledge points in Lark
You: #KnowledgePoint OpenClaw can extend functionality through Skills
OpenClaw: Added to knowledge graph
- Entities: OpenClaw, Skills
- Relationship: Extends functionality
- Category: Technology/Tool

You: #KnowledgePoint find-skills can discover new Skills
OpenClaw: Added to knowledge graph
- Entities: find-skills, Skills
- Relationship: Discovers
- Association: OpenClaw → Skills → find-skills

# Analyze relationships between knowledge points
openclaw agent --message "Analyze my knowledge base and extract relationships between knowledge points"

# OpenClaw automatically analyzes:
Discovered relationships:
1. OpenClaw → Includes → Skills (Hierarchical relationship)
2. Skills → Depends on → API Configuration (Dependency relationship)
3. find-skills → Similar to → ProactiveAgent (Similarity relationship)
4. Automation → Applies to → OpenClaw (Application relationship)
5. Docker → Deployment Method → OpenClaw (Method relationship)

Knowledge graph updated

{
  "relationships": [
    {
      "from": "OpenClaw",
      "to": "Skills",
      "type": "Includes",
      "weight": 1.0
    },
    {
      "from": "Skills",
      "to": "Automation",
      "type": "Implements",
      "weight": 0.9
    },
    {
      "from": "find-skills",
      "to": "ProactiveAgent",
      "type": "Used in conjunction with",
      "weight": 0.8
    }
  ]
}

# Generate visual graph
clawhub install knowledge-graph-visualizer
openclaw skills run knowledge-graph-visualizer \
  --input ~/.openclaw/knowledge \
  --output ~/.openclaw/knowledge/graph.html \
  --style "force-directed"

# Open in browser
open ~/.openclaw/knowledge/graph.html

         [AI Tools]
            ↓
        [OpenClaw] ←→ [Automation]
         ↙  ↓  ↘
    [Skills] [Gateway] [Channels]
       ↓
  [find-skills]
  [ProactiveAgent]
  [brave-search]

# Search knowledge points
openclaw agent --message "Search: How to use Skills"

# OpenClaw returns based on the knowledge graph:
Found related knowledge:
1. Skills Basic Concepts (Relevance: 95%)
   - What are Skills
   - Role of Skills
   - How to install Skills

2. Skills Practical Cases (Relevance: 90%)
   - How to use find-skills
   - ProactiveAgent configuration
   - Skills combination application

3. Related Concepts (Relevance: 80%)
   - OpenClaw architecture
   - API configuration
   - Automation workflow

Recommended learning path:
Basic Concepts → Installation & Configuration → Practical Cases → Advanced Applications

openclaw agent --message "Based on my knowledge graph, recommend next learning content."

# OpenClaw analyzes the knowledge graph:
You have mastered:
- OpenClaw basics (100%)
- Skills installation (100%)
- Docker deployment (80%)

Recommended learning:
1. Skills Combination Applications (Priority: High)
   Reason: You have mastered individual Skills and can now learn combined applications.

2. Advanced Automation Workflows (Priority: High)
   Reason: Based on your Skills knowledge, you can build workflows.

3. Knowledge Graph Construction (Priority: Medium)
   Reason: Advanced content, requires more practical experience.

Estimated learning time:
- Skills combination: 2 hours
- Automation workflow: 4 hours
- Knowledge graph: 3 hours

# Weekly knowledge review
openclaw agent --message "Generate this week's knowledge review report"

# OpenClaw generates report:
📊 This Week's Knowledge Review 2026-02-11

📚 New knowledge points: 15
- Docker deployment (5)
- Skills application (7)
- Automation workflow (3)

🔗 New relationships: 23
- Hierarchical relationships: 8
- Dependency relationships: 10
- Application relationships: 5

📈 Knowledge Growth
- Total knowledge points: 156 (+15)
- Knowledge density: Increased by 12%
- Knowledge depth: Increased by 8%

💡 Knowledge Insights
1. Docker-related knowledge forms a complete system
2. Skills application knowledge is growing rapidly
3. Automation workflows require more practice

🎯 Next Week's Plan
1. Supplement knowledge on Skills combination applications
2. Practice automation workflows
3. Build knowledge graph visualization

{
  "domain": "Python Asynchronous Programming",
  "core_concepts": [
    "Coroutine",
    "Event Loop",
    "async/await",
    "Concurrency",
    "Asynchronous IO"
  ],
  "related_concepts": [
    "Multithreading",
    "Multiprocessing",
    "GIL",
    "Performance Optimization"
  ]
}

# Automatically collect resources
openclaw agent --message "Collect learning resources for Python asynchronous programming"

# OpenClaw collects:
- Official documentation: 5 articles
- Technical articles: 20
- Video tutorials: 8
- Open source projects: 15
- Practical cases: 10

Added to knowledge graph

Python Asynchronous Programming
  ├─ Basic Concepts
  │   ├─ Coroutine
  │   ├─ Event Loop
  │   └─ async/await syntax
  ├─ Core Libraries
  │   ├─ asyncio basic usage
  │   ├─ aiohttp network requests
  │   └─ aiofiles file operations
  ├─ Application Scenarios
  │   ├─ Asynchronous web crawler
  │   ├─ Asynchronous API
  │   └─ Asynchronous task queue
  └─ Best Practices
      ├─ Error Handling
      ├─ Performance Optimization
      └─ Debugging Techniques

openclaw agent --message "Based on the knowledge graph, generate a Python asynchronous programming learning path."

# OpenClaw generates:
📚 Learning Path (Estimated 20 hours)

Stage 1: Basic Concepts (4 hours)
- Understand coroutine principles
- Learn async/await syntax
- Master event loop mechanism

Stage 2: Core Libraries (6 hours)
- asyncio basic usage
- aiohttp network requests
- aiofiles file operations

Stage 3: Practical Projects (8 hours)
- Project 1: Asynchronous web crawler (3 hours)
- Project 2: Asynchronous API service (3 hours)
- Project 3: Asynchronous task queue (2 hours)

Stage 4: Advanced Optimization (2 hours)
- Performance optimization techniques
- Error handling best practices
- Debugging tool usage

After each stage is completed, update the knowledge graph.

# Update knowledge graph weekly
openclaw agent --message "Update knowledge graph"

# OpenClaw automatically:
1. Scans new notes
2. Extracts new knowledge points
3. Analyzes new relationships
4. Updates graph structure
5. Generates update report

# Check knowledge graph quality
openclaw agent --message "Check knowledge graph quality"

# OpenClaw analyzes:
Quality report:
- Isolated nodes: 3 (need to establish connections)
- Weak relationships: 5 (need to be strengthened)
- Duplicate nodes: 2 (need to be merged)
- Missing relationships: 8 (need to be supplemented)

Optimization suggestions:
1. Merge duplicate nodes: "OpenClaw Skills" and "Skills"
2. Establish missing relationship: "Docker" → "Deployment" → "OpenClaw"
3. Strengthen weak relationships: Add practical cases
#### Knowledge Export

```bash
# 导出知识图谱
openclaw agent --message "请使用 knowledge-exporter skill 导出知识图谱为 markdown 格式，保存到 ~/knowledge-base.md"

# 生成Markdown格式的知识库
# 可以发布为个人博客或文档

{
  "tracking": {
    "time_tracking": {
      "enabled": true,
      "auto_detect": true,
      "categories": {
        "Work": ["Programming", "Writing", "Meetings"],
        "Learning": ["Reading", "Videos", "Practice"],
        "Rest": ["Rest", "Exercise", "Entertainment"]
      }
    },
    "task_tracking": {
      "enabled": true,
      "track_quality": true,
      "track_difficulty": true
    },
    "automation_tracking": {
      "enabled": true,
      "track_savings": true,
      "track_roi": true
    }
  },
  "reporting": {
    "daily": "20:00",
    "weekly": "Sunday 18:00",
    "monthly": "Last day 18:00"
  }
}

# 查看实时效率数据
openclaw agent --message "显示我的效率仪表盘"

# OpenClaw generated:
📊 Efficiency Dashboard 2026-02-11 15:30

⏰ Today's Time Allocation
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Work ████████████░░░░░░░░ 6.5h (Goal 8h)
Learning ████░░░░░░░░░░░░░░░░ 2h   (Goal 2h)
Rest ██░░░░░░░░░░░░░░░░░░ 1h   (Goal 1.5h)

✅ Task Completion Status
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Planned Tasks: 8
Completed: 6 ✓
In Progress: 1 ⟳
Not Started: 1 ○
Completion Rate: 75% (Goal 80%)

🎯 Focus Analysis
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Deep Work: 4h
Pomodoros: 8
Interruptions: 3
Focus Level: 85% ↑

🤖 Automation Effect
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Automated Tasks: 12
Time Saved: 2.5h
Automation Rate: 60%
ROI: 1200%

💡 Real-time Suggestions
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
1. 1.5h of work time remaining to reach the goal
2. Suggest reducing meetings and increasing deep work
3. Tasks that can be automated: email replies, data organization

Scheme A: Traditional Workflow
- Manually collect information
- Manually organize notes
- Manually publish content

Scheme B: Automated Workflow
- Automatically collect information
- Automatically organize notes
- Automatically publish content

Test Period: 2 weeks
Test Metrics: Time, Quality, Satisfaction

# 启动A/B测试
openclaw agent --message "启动工作流A/B测试"

# Week 1: Use Scheme A
openclaw config set workflow "manual"

# Week 2: Use Scheme B
openclaw config set workflow "automated"

# Test ends, generate comparison report
openclaw agent --message "生成A/B测试报告"

# OpenClaw generated:
📊 A/B Test Report

Scheme A (Traditional Workflow)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Average Time Spent: 8.5h/day
Tasks Completed: 6/day
Content Quality: 85 points
Satisfaction: 70%

Scheme B (Automated Workflow)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Average Time Spent: 3.2h/day ↓62%
Tasks Completed: 8/day ↑33%
Content Quality: 87 points ↑2%
Satisfaction: 90% ↑20%

Conclusion: Scheme B is significantly better than Scheme A
Recommendation: Fully adopt automated workflow

Plan
  ↓
Do
  ↓
Check
  ↓
Act
  ↓
Plan (New Plan)...

openclaw agent --message "制定本周效率优化计划"

# OpenClaw generated:
📋 This Week's Optimization Plan

Goals:
1. Increase work duration to 8h/day
2. Increase task completion rate to 90%
3. Increase automation rate to 70%

Actions:
1. Reduce meeting time (2h → 1h)
2. Increase deep work time (4h → 6h)
3. Automate email replies and data organization

Expected Results:
- Time Savings: 2h/day
- Efficiency Improvement: 30%

# Record daily execution status
openclaw agent --message "记录今日执行情况"

# OpenClaw automatically tracks:
- Work Duration: 7.5h ✓
- Tasks Completed: 8/9 (89%) ✓
- Automation Rate: 65% △

openclaw agent --message "生成本周执行报告"

# OpenClaw analysis:
📊 This Week's Execution Report

Goal Achievement Status:
1. Work Duration: 7.2h/day (Goal 8h) △
2. Task Completion Rate: 87% (Goal 90%) △
3. Automation Rate: 65% (Goal 70%) △

Problem Analysis:
1. Meeting time is still excessive (1.5h)
2. Frequent interruptions (5 times/day)
3. Some tasks were not automated

Improvement Directions:
1. Further reduce meetings
2. Set focus periods
3. Identify automatable tasks

openclaw agent --message "制定改进措施"

# OpenClaw suggests:
💡 Improvement Measures

Measure 1: Meeting Optimization
- Decline unnecessary meetings
- Schedule meetings for the afternoon
- Prepare agenda before meetings

Measure 2: Focus Periods
- 9 AM - 12 PM as focus period
- Turn off notifications
- Decline interruptions

Measure 3: Automation Expansion
- Automate email replies
- Automate data organization
- Automate report generation

Continue execution and track results next week

# Analyze time usage
openclaw agent --message "审计我的时间使用"

# OpenClaw analysis:
⏰ Time Audit Report

High-Value Activities (should increase)
- Deep Work: 4h/day
- Learning & Improvement: 2h/day
- Strategic Thinking: 0.5h/day

Low-Value Activities (should decrease)
- Ineffective Meetings: 1.5h/day
- Email Processing: 1h/day
- Social Media: 0.5h/day

Wasted Time (should eliminate)
- Repetitive Tasks: 1h/day
- Waiting Time: 0.5h/day
- Aimless Browsing: 0.5h/day

Optimization Suggestions:
1. Reduce meetings by 50%, saving 0.75h
2. Automate emails, saving 0.5h
3. Eliminate repetitive tasks, saving 1h
Total potential savings: 2.25h/day

# Analyze energy curve
openclaw agent --message "分析我的能量曲线"

# OpenClaw generated:
📈 Energy Curve Analysis

Peak Energy Periods:
- 09:00-11:00 (Optimal)
- 15:00-17:00 (Sub-optimal)

Low Energy Periods:
- 13:00-14:00 (Afternoon)
- 18:00-19:00 (Evening)

Optimization Suggestions:
1. Schedule for peak periods:
   - Complex tasks
   - Creative work
   - Important decisions

2. Schedule for low periods:
   - Simple tasks
   - Routine work
   - Rest and recovery

3. Energy boost:
   - 20-minute nap
   - 5-minute break every 90 minutes
   - Moderate exercise

# Track efficiency habits
openclaw agent --message "追踪我的效率习惯"

# OpenClaw records:
📊 Habit Tracking (30 days)

Good Habits:
✓ Early rising (28/30 days, 93%)
✓ Morning planning (25/30 days, 83%)
✓ Pomodoro Technique (22/30 days, 73%)
✓ Daily review (20/30 days, 67%)

Bad Habits:
✗ Scrolling on phone before bed (18/30 days, 60%)
✗ Distracted during work (15/30 days, 50%)
✗ Procrastinating tasks (10/30 days, 33%)

Improvement Suggestions:
1. Reinforce good habits: Set reminders
2. Eliminate bad habits: Establish alternative behaviors
3. Track progress: Weekly review

OpenClaw: Good morning! Today's weather is clear, temperature 18°C
Today's Schedule:
- 09:00-12:00 Deep Work (Project Development)
- 14:00-15:00 Team Meeting
- 15:30-17:30 Content Creation

Prepared for you:
- Today's news summary (15 items)
- To-do list (8 tasks)
- Calendar reminders (3 items)

You: Check today's plan
OpenClaw:
Priority Order:
🔴 P0: Fix production bug (2h)
🟡 P1: Complete feature development (3h)
🟢 P2: Write technical article (2h)

Suggested Execution Order:
09:00-11:00 Fix bug (Focus Period)
11:00-12:00 Feature development
15:30-17:30 Write article (Creative work)

OpenClaw: Entering focus mode
- Notifications off
- Distractions blocked
- Pomodoro timer started (25 minutes)

[After 25 minutes]
OpenClaw: Pomodoro completed, 5-minute break
Progress: Bug fix 40%

OpenClaw: Morning work summary
- Work Duration: 3h
- Tasks Completed: 2
- Focus Level: 90%

Afternoon Suggestions:
- 14:00 Team meeting (prepare agenda)
- 15:30 Content creation (materials prepared)

You: Start writing a technical article
OpenClaw:
- Relevant materials searched (10 articles)
- Article outline generated
- Code examples prepared
- Image suggestions generated

Start creating!

openclaw agent --message "生成今日复盘"

OpenClaw:
📊 Today's Review 2026-02-11

✅ Completion Status
- Planned Tasks: 8
- Completed Tasks: 8
- Completion Rate: 100% 🎉

⏰ Time Allocation
- Deep Work: 6h
- Meetings: 1h
- Learning: 1h

🎯 Efficiency Data
- Focus Level: 88%
- Automation Rate: 65%
- Time Saved: 2.5h

💡 Tomorrow's Suggestions
1. Continue to maintain high focus
2. Increase automated tasks
3. Optimize meeting time

Well done! Keep up the good work tomorrow!

Requirements → OpenClaw Analysis → Generate Development Plan → 
Invoke Claude Code → Automated Programming → Test Verification → 
Bug Discovery → Automated Fix → Delivery Completion

✅ 24/7 uninterrupted development
✅ Automated testing and fixing
✅ Stable code quality
✅ 10x increase in development efficiency

You: Help me develop a to-do application,
    Requirements:
    - Python Flask backend
    - React frontend
    - SQLite database
    - RESTful API

OpenClaw: Okay, planning development tasks...

📋 【Development Plan】To-Do Application

🎯 Project Structure:
todo-app/
├── backend/          # Flask backend
│   ├── app.py
│   ├── models.py
│   ├── routes.py
│   └── requirements.txt
├── frontend/         # React frontend
│   ├── src/
│   ├── public/
│   └── package.json
└── README.md

⏳ Starting development...

✅ Step 1: Create Backend (Claude Code)
• Generate Flask application structure
• Define data models
• Implement API routes
• Add database operations

✅ Step 2: Create Frontend (Claude Code)
• Initialize React project
• Create components
• Implement API calls
• Add styles

✅ Step 3: Test Verification
• Unit tests: Passed
• Integration tests: Passed
• API tests: Passed

✅ Development complete!

📊 Development Statistics:
• Total time spent: 15 minutes
• Lines of code: 850 lines
• Test coverage: 85%
• Number of Bugs: 0

📁 Project saved to: ~/projects/todo-app
🚀 Start command:
   cd ~/projects/todo-app
   ./start.sh

You: Production environment reported an error, help me fix it
    Error message: TypeError: 'NoneType' object is not subscriptable

OpenClaw: Analyzing error...

🔍 【Bug Analysis】

Error Type: TypeError
Error Location: app.py:45
Reason: Attempted to access attribute of 'user' when it was None

🔧 Fix Plan:
1. Add null check
2. Return user-friendly error message
3. Log error

⏳ Starting fix...

✅ Fix complete!

📝 Changes:
• Added user null check
• Optimized error handling
• Added logging

🧪 Test Results:
• Unit tests: Passed
• Edge case tests: Passed
• Regression tests: Passed

✅ Automatically deployed to production

You: Help me review the code for this PR

OpenClaw: Reviewing code...

📊 【Code Review Report】

✅ Code Quality: 85/100

🔍 Issues Found:
1. Performance issues (2 instances)
   • Database queries in a loop
   • Unindexed queries

2. Security issues (1 instance)
   • SQL injection risk

3. Code style (3 instances)
   • Inconsistent variable naming
   • Missing comments
   • Overly long function

💡 Optimization Suggestions:
1. Use batch queries to reduce database access
2. Use parameterized queries to prevent SQL injection
3. Refactor long functions to improve readability

🔧 Automated Fixes:
• Security issues fixed
• Performance issues optimized
• Code format standardized

✅ Review complete, merge recommended

# 1. Install Coding Agent
clawhub install coding-agent

# 2. Configure Claude Code
openclaw config set coding.tool "claude-code"
openclaw config set coding.model "claude-3-5-sonnet"

# 3. Configure automation
openclaw config set coding.auto-test true
openclaw config set coding.auto-fix true
openclaw config set coding.auto-deploy false

# 4. Configure code style
openclaw config set coding.style "pep8"
openclaw config set coding.linter "pylint"

Weekly Development Tasks:
• New feature development: 2 times × 4 hours = 8 hours
• Bug fixing: 5 times × 2 hours = 10 hours
• Code review: 10 times × 1 hour = 10 hours
• Documentation writing: 1 time × 3 hours = 3 hours
• Total: 31 hours

After using Coding Agent:
• New feature development: 2 times × 15 minutes = 30 minutes
• Bug fixing: 5 times × 5 minutes = 25 minutes
• Code review: 10 times × 3 minutes = 30 minutes
• Documentation writing: 1 time × 10 minutes = 10 minutes
• Total: 95 minutes

Weekly Savings: 29 hours
Monthly Savings: 116 hours ≈ 14.5 workdays

Read https://moltbook.com/skill.md and follow the instructions to join Moltbook

You: Send me today's weather forecast every morning at 7 AM,
    including temperature, whether to bring an umbrella, and clothing suggestions

OpenClaw: Okay, scheduled task set ✅

☀️ Good morning! Today's Weather Broadcast

📍 Beijing
🌡️ Temperature: 15-25°C
☁️ Weather: Sunny to cloudy
💧 Precipitation Probability: 10%

👔 Clothing Suggestions:
• Large temperature difference between morning and evening, recommend wearing a light jacket
• Hot at noon, can prepare short sleeves

☂️ Travel Suggestions:
• No umbrella needed
• Good air quality, suitable for outdoor activities

Have a wonderful day! 🌈

You: Remind me to sleep every night at 10 PM,
    and tell a 3-minute bedtime story

OpenClaw: Okay, set ✅

🌙 Time to sleep!

Tonight's bedtime story: 'The Secret of the Stars'

Once upon a time, a little star lived at the end of the Milky Way...
(Story content)

💤 Good night, sweet dreams!

# Configure scheduled tasks using Hooks
openclaw hooks create \
  --name "morning-weather" \
  --trigger "cron:0 7 * * *" \
  --action "发送今日天气预报"

# Or use a configuration file
cat > ~/.openclaw/hooks/morning-weather.json << EOF
{
  "name": "morning-weather",
  "trigger": {
    "type": "cron",
    "schedule": "0 7 * * *"
  },
  "action": {
    "type": "sendMessage",
    "channel": "feishu",
    "message": "请发送今日天气预报，包括温度、降水、穿衣和出行建议"
  }
}
EOF

You: Help me download the Xiaohongshu Cover Generator Skill

OpenClaw: Downloading...
✅ Installed: xiaohongshu-cover-generator

You: Put the API Key into environment variables

OpenClaw: Okay, configured ✅

You: Generate a Xiaohongshu cover image with the theme: 'Laba Festival Customs in North and South China'

OpenClaw: Generating...
✅ Cover generated!

You: Convert this article into a webpage and generate an accessible link
    [Attachment: article.md]

OpenClaw: Okay, processing...

OpenClaw: ✅ Webpage generated!

📄 Article Title: OpenClaw User Guide
🔗 Access Link: https://your-site.com/article-123
📊 Page Size: 2.3MB
⏱️ Generation Time: 45 seconds

Webpage includes:
• Responsive layout
• Table of contents navigation
• Code highlighting
• Image optimization

@OpenClaw Help me collect and organize the latest information about OpenClaw,
tell me how to configure and use it, and find 10 best practices for me on GitHub or X.

OpenClaw: ✅ Research complete!

📊 OpenClaw Research Report

【Basic Information】
• Project Name: OpenClaw
• Latest Version: 2026.3.2
• GitHub Stars: 60,000+
• Developer: Peter Steinberger

【Configuration Guide】
1. Installation: curl -fsSL https://openclaw.ai/install.sh | bash
2. Initialization: openclaw onboard
3. Configure Models: Supports Claude, GPT-4, Kimi, etc.
4. Select Channel: Lark, Discord, Telegram, etc.

【10 Best Practices】
1. Use cloud deployment (Tencent Cloud/Volcano Engine)
2. Configure scheduled tasks for automation
3. Install common Skills to extend capabilities
4. Use Hooks to automate workflows
5. Configure multiple models for disaster recovery
6. Use Profiles to manage different scenarios
7. Regularly back up configuration files
8. Monitor Token consumption to optimize costs
9. Join the community for the latest information
10. Share experiences to help others

【Recommended Resources】
• Official Documentation: https://docs.openclaw.ai
• GitHub: https://github.com/openclaw/openclaw
• Discord Community: https://discord.gg/openclaw
• Chinese Tutorial: https://github.com/xianyu110/awesome-openclaw-tutorial

【Related Discussions】
• Trending topics on X: #OpenClaw
• Reddit discussion: r/OpenClaw
• Zhihu Column: OpenClaw Practices

You: Save the information just generated as a markdown document,
    in the 'claw' folder on my computer; you need to create this folder

OpenClaw: Okay, saving...
✅ Saved to: ~/claw/openclaw-research.md

# Automatically collect the latest OpenClaw information every morning at 9 AM
openclaw hooks create \
  --name "daily-openclaw-news" \
  --trigger "cron:0 9 * * *" \
  --action "搜集OpenClaw最新信息并发送到飞书"

You: Help me install opencode

OpenClaw: Installing...
✅ OpenCode installed

You: Create a new directory for the project

OpenClaw: ✅ Created: ~/myopencode

You: Change directory and start opencode

OpenClaw: Starting OpenCode...
✅ OpenCode started

You: Use opencode to develop a Snake game

OpenClaw: Okay, developing...

📋 Development Plan:
• Use HTML5 Canvas
• Implement game logic
• Add scoring system
• Responsive design

⏳ Developing...
✅ Game development complete!

📊 Development Statistics:
• Lines of code: 400 lines
• Development time: 5 minutes
• Number of files: 3 (HTML/CSS/JS)

You: Provide my GitHub information for connection

OpenClaw: Please provide GitHub Token...

You: [Provide Token]

OpenClaw: ✅ GitHub connected

OpenClaw: Pushing to GitHub...
✅ Code pushed!

📦 Repository Information:
• Repository Name: snake-game
• Address: https://github.com/username/snake-game
• Commit Message: Initial commit: Snake game

You: Deploy to Vercel to get an accessible webpage link

OpenClaw: Deploying...
✅ Deployment successful!

🚀 Deployment Information:
• Access Address: https://myopencode.vercel.app
• Deployment Time: 10 seconds
• Status: Running

You: This process is amazing, I plan to write an article,
    help me generate an article directly based on your previous steps,
    with the title: Agent Coding with OpenClaw+OpenCode+GitHub+Vercel

OpenClaw: Okay, generating article...
✅ Article generated and pushed to GitHub!

User Requirements
    ↓
OpenClaw Analysis
    ↓
OpenCode Development
    ↓
Code Generation Complete
    ↓
Push to GitHub
    ↓
Automated Vercel Deployment
    ↓
Website Live
    ↓
Automated Documentation Generation