AI Digest Plugin

Digest AI/tech articles and content into structured learning documents for quick reference and practical application.

Process Flow

flowchart TD
    A([User Trigger]) -->|"/ai-digest URL or text"| B{Input Type}
    B -->|URL| C[Fetch Article via WebFetch]
    B -->|Pasted Text| D[Use Content Directly]
    C --> E[Extract Key Content]
    D --> E
    E --> F{Focus Specified?}
    F -->|Yes e.g. 'breaking changes only'| G[Apply Focused Extraction]
    F -->|No| H[Full Article Analysis]
    G & H --> I[Structure into Template]

    subgraph Template["Document Template"]
        T1[Summary 요약]
        T2[Key Changes / Concepts]
        T3[Practical Applications 실무 적용]
        T4[Code Examples 코드 예제]
        T5[Limitations 주의사항]
        T6[References + Next Steps]
    end

    I --> Template
    Template --> J[Save to learning_repo/digests/YYYY-MM-DD-ai-topic.md]
    J --> K{auto_commit?}
    K -->|Yes| L[Git Commit & Push]
    K -->|No| M([Done ✅])
    L --> M

Overview

With the rapid pace of AI development, keeping track of new features, changes, and best practices is challenging. The AI Digest plugin helps you capture and organize AI/tech content from articles, blog posts, and announcements into structured, searchable learning documents.

Features

• URL Analysis: Fetch and analyze articles directly from URLs
• Direct Content: Paste article text for immediate analysis
• Focused Extraction: Specify what to focus on (e.g., “breaking changes only”)
• AI-Optimized Template: Structured format for quick reference
• Bilingual Support: Korean/English documentation
• Git Integration: Auto-commit and push (optional)
• Shared Config: Reuses learning-summary configuration

Installation

For Users: Install via npx

# Install from repository
npx github:JayKim88/claude-ai-engineering ai-digest

For Developers: Local Development

# Clone and link
cd ~/Documents/Projects
git clone https://github.com/JayKim88/claude-ai-engineering.git
cd claude-ai-engineering
npm run link

Usage

Basic Usage

# Digest article from URL
/ai-digest https://anthropic.com/news/claude-sonnet-4-5

# Analyze with specific focus
/ai-digest "Focus on API changes" https://docs.anthropic.com/...

# Digest pasted content
/ai-digest "Analyze this: [paste your article text]"

Alternative Trigger Phrases

# English
"digest this article" + URL
"analyze this AI news" + URL
"summarize this tech blog" + URL

# Korean
"이 글 정리해줘" + URL
"AI 뉴스 요약" + URL
"기술 블로그 분석" + URL

Document Structure

Generated documents include:

# Article Title

> **Source**: [URL]
> **Date**: 2026-01-25
> **Tags**: #ai #topic

## 요약 (Summary)
Brief overview in Korean

## 주요 변경사항 / 새로운 개념 (Key Changes/Concepts)
- What changed or what's new
- Why it matters
- Impact on your work

## 실무 적용 방법 (Practical Applications)
How to use this in real projects

## 코드 예제 (Code Examples)
Runnable code snippets with explanations

## Before/After 비교 (if applicable)
Migration guides and comparisons

## 주의사항 / 제한사항 (Limitations)
Important warnings and gotchas

## 참고 링크 (References)
Original URL and related resources

## Next Steps
Follow-up actions and topics to explore

Configuration

The plugin reuses the learning-summary configuration file.

Location: ~/.claude/skills/learning-summary/config.yaml

# Dedicated AI learning repository
learning_repo: "/Users/username/Documents/Projects/ai-learning"

# Auto-commit to git
auto_commit: false

# Auto-push to remote (requires auto_commit: true)
auto_push: false

Note: ai-digest always saves to the digests/ subfolder (not learnings/) to separate article digests from conversation summaries.

If config doesn’t exist, you’ll be prompted to provide the learning repository path.

Examples

Example 1: Digest Claude Release Notes

/ai-digest https://anthropic.com/news/claude-sonnet-4-5

Output: /Users/username/Documents/Projects/ai-learning/digests/2026-01-25-ai-claude-sonnet-4-5-release.md

Captures:

• New features and improvements
• Performance benchmarks
• Breaking changes and migrations
• Code examples for new capabilities
• Practical use cases

Example 2: Focus on Specific Aspects

/ai-digest "Focus on breaking changes and migration" https://openai.com/blog/gpt-5

Result: Document emphasizes:

• API breaking changes
• Deprecated features
• Migration steps with code examples
• Compatibility notes

Example 3: Batch Processing

# Process multiple articles in one session
/ai-digest https://anthropic.com/news/article-1
/ai-digest https://openai.com/blog/article-2
/ai-digest https://langchain.com/blog/article-3

Each article is saved as a separate document in your ai-learning repository.

Output Location

Documents are saved to:

{learning_repo}/digests/YYYY-MM-DD-ai-[topic].md

Example:

/Users/jaykim/Documents/Projects/ai-learning/
├── learnings/                    # conversation summaries (learning-summary)
└── digests/                      # article digests (ai-digest)
    ├── 2026-01-25-ai-claude-sonnet-4-5-release.md
    ├── 2026-01-25-ai-openai-gpt5-features.md
    └── 2026-01-25-ai-langchain-updates.md

Use Cases

1. Daily AI News Digest

Stay up-to-date with rapid AI developments:

# Morning routine
/ai-digest https://anthropic.com/news/latest
/ai-digest https://openai.com/blog/recent

2. Project-Specific Research

Capture relevant information for current projects:

/ai-digest "Focus on production deployment" https://docs.anthropic.com/production

3. Learning New Concepts

Build a personal knowledge base:

/ai-digest https://blog.example.com/understanding-transformers
/ai-digest https://blog.example.com/rag-patterns

Tips

1. Use Descriptive Focus: Specify what you want to extract

   • “Focus on breaking changes”
   • “Extract code examples only”
   • “Summarize key features”

2. Batch Similar Topics: Process related articles together for better context

3. Enable Auto-commit: Set auto_commit: true for automatic version control

4. Tag Consistently: Documents include tags for easier searching later

5. Review Generated Docs: Edit documents to add personal notes in the “메모” section

Differences from learning-summary

Feature	ai-digest	learning-summary
Input	URL or article text	Current conversation
Focus	AI/tech content	General learning
Output Folder	digests/	learnings/
Template	AI-specific sections	Generic learning sections
Use Case	Rapid info capture	Session summary
When to Use	New articles, updates	End of conversation

Together: Use both plugins for comprehensive learning documentation:

• ai-digest: Capture external content
• learning-summary: Document conversations

Troubleshooting

Issue: URL fetch fails

Solution: Paste the article content directly:

/ai-digest "Analyze this: [paste full article text]"

Issue: Config not found

Solution: Create config file or provide path when prompted:

# ~/.claude/skills/learning-summary/config.yaml
learning_repo: "/path/to/your/ai-learning"
auto_commit: false
auto_push: false

Issue: Document not saved

Solution: Check directory permissions:

ls -la /path/to/ai-learning/digests

Issue: Git commit fails

Cause: Not a git repository

Solution: Initialize git:

cd /path/to/ai-learning
git init
git remote add origin <your-repo-url>

Workflow Integration

Daily Learning Routine

# 1. Collect AI news URLs during the day
# 2. At end of day, batch process:
/ai-digest https://url1
/ai-digest https://url2
/ai-digest https://url3

# 3. Review generated documents
cd ~/Documents/Projects/ai-learning/digests
ls -lt | head -10  # See recent docs

# 4. Commit manually or let auto_commit handle it
git add .
git commit -m "Daily AI updates: 2026-01-25"
git push

Project Research

# Research specific topic
/ai-digest "Focus on implementation details" https://blog.example.com/rag-systems
/ai-digest "Extract performance metrics" https://blog.example.com/rag-benchmarks

# Document your experiments
# (Use learning-summary for conversation notes)

Related Plugins

• learning-summary: Document conversation learnings
• project-insight: Analyze codebase structure and quality

Roadmap

• Batch URL processing (multiple URLs in one command)
• Update existing documents (incremental updates)
• Search existing learnings before creating duplicates
• Export to Notion/Obsidian
• Tag-based organization
• Weekly digest summaries

Contributing

Contributions welcome! See the main repository for guidelines.

Support

• Issues: https://github.com/JayKim88/claude-ai-engineering/issues
• Discussions: https://github.com/JayKim88/claude-ai-engineering/discussions

License

MIT License

Author

Jay Kim

• GitHub: @JayKim88

---

Stay current with AI developments. Capture knowledge efficiently. Build your learning repository.