openclaw-workspace

Summary

This workspace provides resources, modular agent skills, and tools for developing, running, and customizing AI agents and related projects using Python within the OpenClaw ecosystem. It includes documentation, skill modules, agent memory mechanisms, and example projects for news aggregation, daily AI content websites, and email handling.

Included Assets

• Core Instructions

  • AGENTS.md — agent workspace usage and memory handling
  • SOUL.md — agent identity, behavior, and operational boundaries
  • MEMORY.md — structural and long-term memory instructions (in Chinese)
  • TOOLS.md — guidance for recording local environment details

• Featured Projects & Examples

  • agent-skills-project/ — modular agent skill orchestration system (Python)
  • ai-daily-website/ — automated AI news & paper aggregation website
  • ai-news-site/ — real-time AI news aggregation platform with RESTful APIs
  • python-agent-demo/ — Python agent skills demo with workflow and testing

• Skills & Third-party Libraries

  • skills/imap-smtp-email/ — skill for reading/sending email via IMAP/SMTP, with supporting Node.js packages

• Documentation & Configuration

  • Project-level and skill-level README files
  • Example configuration files (e.g., .env for email skill, config.yaml, requirements.txt)

How to Use

1. Review Agent Documentation

   • Start with AGENTS.md for agent lifecycle and memory instructions.
   • Read SOUL.md to understand the intended personality, responsibility, and session boundaries.
   • Use MEMORY.md for long-term memory guidelines and examples (especially for Chinese users).
   • Consult TOOLS.md for organizing local, user-specific environment notes.

2. Run Example Projects

   • Python projects (e.g., agent-skills-project, python-agent-demo):
     • Install Python dependencies:

       pip install -r requirements.txt
       

     • Run demos or workflows as described in the respective README.
   • Web projects (e.g., ai-daily-website):
     • Follow setup instructions in project README (npm install or python3 -m http.server as needed).
   • News sites:
     • Use shell scripts to start/stop server or deploy content per documentation.

3. Explore and Configure Skills

   • For skills such as imap-smtp-email, edit the provided .env file with your credentials.
   • Install Node.js dependencies with npm install in the skill's directory.
   • Use provided scripts/commands to test connectivity and perform email operations.

4. Extend and Customize

   • Develop new skills or agents by following the base class structures and examples in the codebase.
   • Document any environment-specific configurations in TOOLS.md.

Notes

• Each project or skill is self-contained; refer to its README.md for specific setup and usage.
• The workspace emphasizes modularity: agent skills can be reused or modified without affecting core memory or local notes.
• Long-term agent memories and session continuity are designed to be manually curated and regularly reviewed.
• Sensitive credentials and personal data should be managed securely (e.g., use .env files, follow privacy guidelines in documentation).
• Some assets and documentation are in Chinese to support relevant user workflows.
• Node.js is required for certain skills; Python 3 is recommended for core agent projects.
• File paths and usage examples conform to standard OpenClaw workspace conventions.