🦞 ClawLearnFeed

ClawLearnFeed is a personal learning feed skill for OpenClaw/OpenKrab ecosystem. It curates and summarizes the latest content in AI/ML, LLM agents, data engineering, algorithmic trading, RAG systems, and Thailand-specific topics using only free sources and local processing.

Features

Personal Learning Feed: Curates content tailored to your interests from your GitHub profile and specified topics
Free Sources Only: Uses RSS feeds, YouTube RSS, and X/Twitter (no paid APIs required)
Local Processing: All summarization done with local LLM (Ollama free models)
Daily Briefings: Automated Telegram/Discord notifications with 3-5 curated items
ClawMemory Integration: Vector storage for historical search and retrieval
Self-Improvement: Learns from your feedback to improve content relevance
Thailand Focus: Specialized content for Thai developers and AI practitioners

Learning Flow

flowchart TD
    A([Daily Cron Trigger]) --> B[collect_feeds.py]
    B --> C{Content Sources}
    C -->|RSS Feeds| D[Parse XML Feeds]
    C -->|YouTube| E[YouTube RSS Channels]
    C -->|X/Twitter| F[X Keyword Search]
    D --> G[summarize.py]
    E --> G
    F --> G
    G --> H{LLM Relevance Filter}
    H -->|Score > 0.7| I[Store in ClawMemory]
    H -->|Score < 0.7| J[Discard]
    I --> K[briefing.py]
    K --> L[Format Daily Digest]
    L --> M[Send Telegram/Discord]
    M --> N[feedback.py]
    N --> O[User Feedback Loop]
    O --> P[Update Topics/Filter]
    P --> A

Getting Started

Prerequisites

OpenClaw CLI (recommended)
Python 3.8+ (for advanced features)
Ollama (free local LLM for summarization)
Optional: Telegram/Discord webhook for notifications

Installation

Via ClawFlow (Recommended)

clawflow install claw-learnfeed

Manual Installation

git clone https://github.com/openkrab/claw-learnfeed.git ~/.openclaw/skills/claw-learnfeed
mkdir -p ~/.openclaw/workspace/.learnings

Quick Test

# Test basic functionality
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --test

# Test summarization
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --test

# Generate sample briefing
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --preview

Core Commands

Feed Collection

# Manual feed collection
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --run

# Test specific sources
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --source rss --test

# Check feed health
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --health

Content Summarization

# Summarize collected feeds
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --run

# Test LLM connectivity
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --test-ollama

# Adjust summary length
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --length short

Daily Briefing

# Generate and send briefing
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --send

# Preview briefing without sending
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --preview

# Test webhook connection
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --test-webhook

Feedback Processing

# Process user feedback
python ~/.openclaw/skills/claw-learnfeed/scripts/feedback.py --process

# View feedback history
python ~/.openclaw/skills/claw-learnfeed/scripts/feedback.py --history

# Update topic weights
python ~/.openclaw/skills/claw-learnfeed/scripts/feedback.py --update-weights

Integration Commands

ClawMemory Storage

# Store summaries in vector DB
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --store-memory

# Search historical content
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --query "LLM agents Thailand"

# Export memory for backup
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --export-memory

ClawFlow Automation

# Install via ClawFlow
clawflow install claw-learnfeed

# Schedule daily briefing
clawflow schedule claw-learnfeed --cron "0 8 * * *" --command briefing

# Update to latest version
clawflow update claw-learnfeed

Telegram/Discord Setup

# Configure Telegram webhook
curl -X POST https://api.telegram.org/bot<TOKEN>/setWebhook?url=<WEBHOOK_URL>

# Configure Discord webhook
curl -X POST <DISCORD_WEBHOOK_URL> -H "Content-Type: application/json" -d '{"content":"Test"}'

# Test notification
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --test-notification

Project Structure

claw-learnfeed/
├── README.md              # This file
├── SKILL.md               # Skill specification and API
├── ClawFlow.yaml          # Installation metadata
├── config.yaml            # Topic and source configuration
├── feeds/                 # Free RSS feed definitions
│   ├── ai_ml_feeds.txt    # AI/ML RSS sources
│   ├── trading_feeds.txt  # Trading and finance feeds
│   └── thai_feeds.txt     # Thailand-specific feeds
├── scripts/               # Core functionality
│   ├── collect_feeds.py   # RSS/X/YouTube aggregation
│   ├── summarize.py       # LLM summarization
│   ├── briefing.py        # Daily digest formatting
│   └── feedback.py        # Self-improvement loop
├── templates/             # Output templates
│   ├── briefing.md        # Daily briefing format
│   └── feedback.md        # Feedback processing
└── .learnings/            # User feedback and improvements
    ├── FEEDBACK.md        # User ratings and comments
    └── TOPICS.md          # Topic adjustment history

Configuration

Topics Setup

# ~/.openclaw/skills/claw-learnfeed/config.yaml
topics:
  - "AI agents"
  - "LLM RAG systems"
  - "algorithmic trading"
  - "data engineering pipelines"
  - "Thailand AI news"
  - "Rust security research"

sources:
  rss_feeds:
    - "https://aiweekly.co/rss"
    - "https://ai.googleblog.com/atom.xml"
  youtube_channels:
    - "UCbfYPyITQ-7l4upoX8nvctg"  # Two Minute Papers
  x_keywords:
    - "AI Thailand"
    - "LLM agents"

ollama:
  model: "llama3:8b"
  endpoint: "http://localhost:11434"

notifications:
  telegram_webhook: "https://api.telegram.org/bot..."
  discord_webhook: "https://discord.com/api/webhooks/..."

Auto Topic Discovery

# Analyze GitHub profile for topics
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --analyze-github @JonusNattapong

# Analyze X/Twitter posts
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --analyze-twitter @JonusNattapong

Testing

# Full system test
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --test-full

# Test individual components
python ~/.openclaw/skills/claw-learnfeed/scripts/summarize.py --test-ollama
python ~/.openclaw/skills/claw-learnfeed/scripts/briefing.py --test-webhook
python ~/.openclaw/skills/claw-learnfeed/scripts/feedback.py --test-processing

# Performance test
python ~/.openclaw/skills/claw-learnfeed/scripts/collect_feeds.py --benchmark

Sample Daily Briefing

🦞 ClawLearnFeed - Daily Digest (2 มี.ค. 2026)

📚 AI Agents
• DeepMind's new multi-agent framework [link]
  สรุป: Framework ใหม่สำหรับ coordinate AI agents ใน complex tasks...

📈 Algorithmic Trading
• LSTM model updates in TensorFlow 2.16 [link]
  สรุป: Performance improvements และ new features สำหรับ trading models...

🇹🇭 Thailand AI News
• Bangkok Post: AI policy update [link]
  สรุป: รัฐบาลประกาศนโยบาย AI ใหม่ มุ่งเน้น digital transformation...

💡 Feedback: /rate 1 good | /rate 1 bad | /skip trading
🔍 Search past: /query "LLM agents Thailand"

Contributing

Fork the repository
Create feature branch: git checkout -b feature-name
Add free RSS feeds: update feeds/*.txt
Test with local Ollama: python scripts/summarize.py --test
Submit pull request

About

Resources

ClawMemory - Vector storage backend
ClawFlow - Automation system
Ollama - Local LLM runtime
Awesome ML AI RSS Feeds - Free RSS sources

Free Sources Included

RSS Feeds: 50+ AI/ML feeds from Google, DeepMind, TensorFlow, etc.
YouTube Channels: Two Minute Papers, Sentdex, 3Blue1Brown, Andrew Ng
X/Twitter: Keyword-based search using OpenClaw tools
Thailand News: Bangkok Post, Thai PBS RSS feeds

Performance Metrics

Daily Items: 3-5 curated summaries per briefing
Processing Time: <5 minutes with local LLM
Storage: <1MB/day in ClawMemory
Accuracy: >80% relevance after feedback training

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
feeds		feeds
scripts		scripts
templates		templates
ClawFlow.yaml		ClawFlow.yaml
README.md		README.md
SKILL.md		SKILL.md
config.yaml		config.yaml

Folders and files

Latest commit

History

Repository files navigation

🦞 ClawLearnFeed

Features

Learning Flow

Getting Started

Prerequisites

Installation

Via ClawFlow (Recommended)

Manual Installation

Quick Test

Core Commands

Feed Collection

Content Summarization

Daily Briefing

Feedback Processing

Integration Commands

ClawMemory Storage

ClawFlow Automation

Telegram/Discord Setup

Project Structure

Configuration

Topics Setup

Auto Topic Discovery

Testing

Sample Daily Briefing

Contributing

About

Resources

Free Sources Included

Performance Metrics

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages